Abstract

The patterns of adaptive radiations in ancient lakes provide valuable clues to mechanisms of speciation and adaptation. In contrast to vertebrate radiations, for instance in fishes or finches, invertebrate species flocks have been largely neglected. While the increase in molecular data narrows this gap, the anatomical basis for interpreting these data against the background of evolutionary hypotheses is still widely lacking. Here we evaluate anatomical findings in the live-bearing pachychilid freshwater gastropod Tylomelania, which has radiated extensively in ancient lakes in the Indonesian island, Sulawesi; we have aimed at reconciling these data with recently obtained molecular phylogenetic evidence. Discovered more than a century ago, the speciose and phenotypically diverse species flock with 34 currently described taxa was only occasionally cited as an example of adaptive radiation in ancient lakes, while anatomical data were entirely lacking. Our study of anatomical characters reveals very low qualitative variation at the species level. Thus, contrary to earlier views we suggest the existence of a single monophyletic lineage endemic to this island. The most conspicuous feature of Tylomelania is its uterine brooding strategy, i.e. retaining eggs and embryos in the pallial oviduct. This is unique among South-East Asian pachychilids. Within the uterine brood pouch the offspring is surrounded by considerable amounts of nutritive material produced by a very large albumin gland, and the embryos are produced continuously. The shelled juveniles of some species are the largest known so far in viviparous gastropods, measuring almost 2 cm in length when hatching. This combination of reproductive features in Tylomelania, characterized by a high amount of maternal investment, is considered to be ovoviviparous, rendering its brooding strategy unique also among other gastropods. In addition, our data reject a previously assumed close relationship to other South-East Asian pachychilids and instead suggest the North Australian Pseudopotamis as sister group to Tylomelania. These findings have significant consequences for the phylogenetic interpretation of morphological characters of Tylomelania in an evolutionary and biogeographical context, leading to the hypothesis that the common ancestor of both genera originated somewhere on the northern Australian continental margin. © 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 85, 513–542.

INTRODUCTION

Adaptive radiations on islands represent fascinating model systems for the study of speciation and phenotypic diversification. Some have even gained textbook status, either due to their sheer extent in terms of species numbers, as for example the African cichlids (e.g. Barlow, 2000; Kornfield & Smith, 2000), or because they have played an important role in the development of modern evolutionary theory, such as the Galapagos finches (e.g. Lack, 1947; Burns, Hackett & Klein, 2002; Grant & Grant, 2002). In recent years the application of new molecular techniques such as PCR or microsatellites have revolutionized the approach to, and understanding, of some well known radiations (see, e.g. Givnish & Sytsma, 1997;,Sturmbauer, 1998; Sherbakov, 1999; Schluter, 2000; Burns et al., 2002; Salzburger et al., 2002). The correlation of phenotypic and genetic diversity has proved especially illuminating (see e.g. Sturmbauer & Meyer, 1992; Meyer, 1993 on Lake Victoria cichlids). Deplorably though, while it has become sufficiently inexpensive and easy to scan large numbers of species or even populations with genetic markers, traditional morphological studies have been neglected at the same time. The latter approach, however, provides the very data needed to successfully use molecular results in developing and testing hypotheses that aim to explain the origin of organismic diversity and disparity. Accordingly, any discussion of adaptation and intralacustrine speciation without a solid morphological data basis is at best incomplete.

Invertebrate radiations, which are much less studied than species flocks in vertebrates, provide the best examples for this effect. The so-called ‘thalassoid’ (i.e. marine-like) gastropod assemblage in ancient Lake Tanganyika is among the best known (see review in Glaubrecht, 1996). It has been repeatedly (albeit erroneously) cited as a model case of intralacustrine diversification of the pantropical family Thiaridae (Boss, 1978; Michel, 2000; West & Michel, 2000; West et al., 2003) and compared to other presumed thiarid radiations, for example in Lake Biwa (Japan) and Lake Malawi (Michel, 1994). Recent anatomical studies have revealed, however, that all species in Lake Tanganyika previously considered as Thiaridae instead belong to the Paludomidae (Glaubrecht, 1999; Strong & Glaubrecht, 2002, 2003; Glaubrecht & Strong, 2004). This finding, which has recently been supported by molecular data (Wilson, Glaubrecht & Meyer, 2004), renders previous taxonomy-dependent speculations − whether there is, for example, an intrinsic proneness to radiate in ‘thiarids’ (Michel, 1994) − superfluous. In addition, detailed studies of individual taxa within this species flock, such as Tanganyicia (Strong & Glaubrecht, 2002) and Stanleya (Strong & Glaubrecht, 2003) have revealed a far greater diversity of reproductive modes than assumed hitherto. The implications of these new data for interpretation of the adaptive radiation of the Lake Tanganyika gastropods are wide-ranging, given that viviparity has long been considered one of the decisive factors in molluscan adaptive radiations in ancient lakes (Cohen & Johnston, 1987; Johnston & Cohen, 1987; Michel, 1994).

Consequently, in this paper we aim to use a traditional anatomical approach to investigate assumptions about a species flock of live-bearing freshwater gastropods (Caenogastropoda: Cerithioidea: Pachychilidae) in ancient lakes on the Indonesian island of Sulawesi (Rintelen et al., 2004) in order to reconcile morphological data with the most recent molecular phylogenetic data. This radiation was discovered by the Swiss naturalists Fritz and Paul Sarasin, who systematically explored most of the island during research expeditions in 1894–95 and 1902–03, focusing on geology, anthropology and zoology (Sarasin & Sarasin, 1905). They described 16 new gastropods endemic to the central lakes and their drainages (Sarasin & Sarasin, 1897, 1898; Table 1). Later additions (Kruimel, 1913) increased this number by eight, leading to a total of 24 described snail species. Following the recent description of three new taxa (Rintelen & Glaubrecht, 2003), and also including riverine taxa, the total number of pachychilid species currently known from Sulawesi stands at 34 (see Table 1).

Table 1.

Pachychilid species endemic to Sulawesi and their distribution. All species which have not been originally assigned to Tylomelania (author names in brackets) have been described as ‘Melania’ and were later transferred to Brotia (see Introduction and Discussion for the reasoning behind their transfer to Tylomelania). Asterisks indicate taxa originally described as subspecies. The distribution of species endemic to the Malili lakes has not been further specified due to taxonomic uncertainty at the species level. However, every species in the Malili system is restricted to a single lake or subset of lakes

Species Distribution (lake system or region) 
Tylomelania abendanoni (Kruimel, 1913) Malili lakes 
T. bakara Rintelen & Glaubrecht, 2003 Malili lakes 
T. carbo Sarasin & Sarasin, 1897 Lake Poso 
T. carota (Sarasin & Sarasin, 1898) Kalaena drainage 
T. celebicola (Sarasin & Sarasin, 1898)* Central Sulawesi 
T. centaurus (Sarasin & Sarasin, 1898) Lake Poso 
T. connectens Sarasin & Sarasin, 1898* Poso River 
T. gemmifera (Sarasin & Sarasin, 1897) Malili lakes 
T. helmuti Rintelen & Glaubrecht, 2003 Malili drainage 
T. insulaesacrae (Sarasin & Sarasin, 1897) Malili lakes 
T. kruimeli Rintelen & Glaubrecht, 2003 Malili lakes 
T. kuli (Sarasin & Sarasin, 1898) Lake Poso 
T. lalemae (Kruimel, 1913) Malili lakes 
T. mahalonensis (Kruimel, 1913) Malili lakes 
T. mahalonica (Kruimel, 1913) Malili lakes 
T. masapensis (Kruimel, 1913) Malili lakes 
T. molesta (Sarasin & Sarasin, 1897) Malili lakes 
T. monacha (Sarasin & Sarasin, 1899) Malili lakes 
T. neritiformis (Sarasin & Sarasin, 1897) Poso River 
T. palicolarum (Sarasin & Sarasin, 1897) Malili lakes 
T. patriarchalis (Sarasin & Sarasin, 1897) Malili lakes 
T. perconica (Sarasin & Sarasin, 1898)* Palopo plain 
T. perfecta (Mousson, 1849) south, south-east and central Sulawesi 
T. porcellanica Sarasin & Sarasin, 1897 Poso River 
T. robusta (Martens, 1897) Toraja region 
T. sarasinorum (Kruimel, 1913) Malili lakes 
T. scalariopsis (Sarasin & Sarasin, 1897) Lake Poso drainage 
T. tominganensis (Kruimel, 1913) Malili lakes 
T. tomoriensis (Sarasin & Sarasin, 1898) Tomori area 
T. toradjarum (Sarasin & Sarasin, 1897) Lake Poso 
T. towutensis (Sarasin & Sarasin, 1897)* Malili lakes 
T. towutica (Kruimel, 1913) Malili lakes 
T. wallacei (Reeve, 1860) Maros karst 
T. zeamais (Sarasin & Sarasin, 1897) Malili lakes 
Species Distribution (lake system or region) 
Tylomelania abendanoni (Kruimel, 1913) Malili lakes 
T. bakara Rintelen & Glaubrecht, 2003 Malili lakes 
T. carbo Sarasin & Sarasin, 1897 Lake Poso 
T. carota (Sarasin & Sarasin, 1898) Kalaena drainage 
T. celebicola (Sarasin & Sarasin, 1898)* Central Sulawesi 
T. centaurus (Sarasin & Sarasin, 1898) Lake Poso 
T. connectens Sarasin & Sarasin, 1898* Poso River 
T. gemmifera (Sarasin & Sarasin, 1897) Malili lakes 
T. helmuti Rintelen & Glaubrecht, 2003 Malili drainage 
T. insulaesacrae (Sarasin & Sarasin, 1897) Malili lakes 
T. kruimeli Rintelen & Glaubrecht, 2003 Malili lakes 
T. kuli (Sarasin & Sarasin, 1898) Lake Poso 
T. lalemae (Kruimel, 1913) Malili lakes 
T. mahalonensis (Kruimel, 1913) Malili lakes 
T. mahalonica (Kruimel, 1913) Malili lakes 
T. masapensis (Kruimel, 1913) Malili lakes 
T. molesta (Sarasin & Sarasin, 1897) Malili lakes 
T. monacha (Sarasin & Sarasin, 1899) Malili lakes 
T. neritiformis (Sarasin & Sarasin, 1897) Poso River 
T. palicolarum (Sarasin & Sarasin, 1897) Malili lakes 
T. patriarchalis (Sarasin & Sarasin, 1897) Malili lakes 
T. perconica (Sarasin & Sarasin, 1898)* Palopo plain 
T. perfecta (Mousson, 1849) south, south-east and central Sulawesi 
T. porcellanica Sarasin & Sarasin, 1897 Poso River 
T. robusta (Martens, 1897) Toraja region 
T. sarasinorum (Kruimel, 1913) Malili lakes 
T. scalariopsis (Sarasin & Sarasin, 1897) Lake Poso drainage 
T. tominganensis (Kruimel, 1913) Malili lakes 
T. tomoriensis (Sarasin & Sarasin, 1898) Tomori area 
T. toradjarum (Sarasin & Sarasin, 1897) Lake Poso 
T. towutensis (Sarasin & Sarasin, 1897)* Malili lakes 
T. towutica (Kruimel, 1913) Malili lakes 
T. wallacei (Reeve, 1860) Maros karst 
T. zeamais (Sarasin & Sarasin, 1897) Malili lakes 

Apart from a brief mention by Sarasin & Sarasin (1898) of some specimens containing embryonic shells, hardly any anatomical data were known, with the exception of operculum and radula descriptions given for all species then known by Sarasin & Sarasin (1897, 1898) and Kruimel (1913). Although the Sulawesi pachychilids were not studied in more detail for the following 80 years, and despite this obvious lack of data, this lacustrine gastropod flock has since been cited in only cursory statements as an example of intralacustrine radiation (e.g. Wesenberg-Lund, 1939; Brooks, 1950; Davis, 1982). Consequently, ideas developed by these authors on the lacustrine evolution in Sulawesi suffer from relying on erroneous taxonomy-based assumptions. For instance, the species flock in the two ancient lake systems of the island (Fig. 1) was thought to comprise two conchologically distinct genera, BrotiaH. Adams, 1866 and Tylomelania Sarasin & Sarasin, 1897 (see Discussion for further details on the taxonomy of the Sulawesi pachychilids). While Tylomelania was originally perceived to be endemic, with four taxa found only in Lake Poso, species of Brotia are widespread not only in Sulawesi but also in South-East Asia.

Figure 1

A, Indonesia and Sulawesi. B, Sulawesi and its lakes. C, Lake Poso. D, Malili lake system.

Figure 1

A, Indonesia and Sulawesi. B, Sulawesi and its lakes. C, Lake Poso. D, Malili lake system.

The assignment of most Sulawesi pachychilids to Brotia implicitly led to two assumptions, which are relevant to evolutionary considerations. First, that there was a close phylogenetic and biogeographical relationship of all South-East Asian congeneric taxa including those on Sulawesi. Second (and deducible from this), that species of Brotia on Sulawesi, in analogy to known viviparity by means of a subhaemocoelic broodpouch (i.e. positioned in the neck region of the head-foot) found in congeneric taxa from the South-East Asian mainland (Solem, 1966; Davis, 1971), have a similar reproductive biology that might facilitate lacustrine radiation.

Recent studies have revealed both assumptions to be invalid. Based on the first modern collections made by Philippe Bouchet (MNHN) in 1991, preliminary evidence of a rather distinct brooding structure utilizing the pallial gonoduct was found in Sulawesi pachychilids (Rintelen & Glaubrecht, 1999). Meanwhile, parallel research on representatives of Brotia from other regions in South-East Asia uncovered still different viviparous strategies and reproductive morphologies (Köhler & Glaubrecht, 2001, 2003). We consequently suggested that the species from Sulawesi formerly assigned to Brotia and Tylomelania actually represent an independent pachychilid lineage distinct from all other South-East Asian taxa (Rintelen & Glaubrecht, 1999; Köhler & Glaubrecht, 2001). In this paper we reassess the evidence for the existence of two distinct genera within this lineage. In our anatomical description we have for formal reasons used Tylomelania for all pachychilid species on Sulawesi, thus including taxa formerly assigned to Brotia. This is followed by a discussion of the systematic and biogeographical affinities of the Sulawesi pachychilids in light of these detailed anatomical data.

MATERIAL AND METHODS

This study is based on material collected in several field-trips to Sulawesi in August 1999, March 2000, September−December 2002 and September−October 2003. Systematic collections were made in rivers and streams throughout the distribution area of the Sulawesi pachychilids and in the ancient lakes as indicated in Table 1 and Figure 13. All material collected in these field-trips was preserved in 70–95% ethanol. Voucher specimens employed in this study, including histological slides and DNA samples, are deposited in the Malacological Department of the Museum of Natural History, Berlin (ZMB). Locality details and both museum and GenBank accession numbers for all sequenced animals are listed in the Appendix. For the compilation of the distributional data, all accessible records from the main sampling expeditions to Sulawesi were evaluated (see caption of Fig. 13).

Figure 13

Tylomelania, distribution. Based on material collected by Sarasin & Sarasin (1893−1896, 1902−1903 NMB), Elbert (Sunda Expedition 1909–1910; SMF), Abendanon (1909−1910; ZMA), NAMRU Expedition (1971–73; ANSP), Bouchet (1991; MNHN) and the authors (1999, 2000, 2002, 2003; ZMB & MZB).

Figure 13

Tylomelania, distribution. Based on material collected by Sarasin & Sarasin (1893−1896, 1902−1903 NMB), Elbert (Sunda Expedition 1909–1910; SMF), Abendanon (1909−1910; ZMA), NAMRU Expedition (1971–73; ANSP), Bouchet (1991; MNHN) and the authors (1999, 2000, 2002, 2003; ZMB & MZB).

Morphology

Shells were measured to 0.1 mm using an electronic calliper. Standard shell parameters were taken following Dillon (1984). Embryonic shells were measured using an ocular micrometer, and their parameters taken as in adult specimens.

Anatomy was studied with a stereo-microscope, and drawings were done with a prisma. Radulae and embryonic shells were studied by scanning electron microscopy (SEM). The radula was cleaned enzymatically with proteinase K as described by Holznagel (1998), sonicated and then mounted on aluminium specimen stubs with adhesive pads. Embryonic shells were  cleaned  mechanically,  sonicated,  and  mounted on adhesive carbon-coated pads. Both radulae and embryonic shells were coated with gold-palladium and studied on a LEO 1450VP scanning electron microscope (software: 32 V02.03) at 10 kV. The dimensions of the initial whorl of embryonic shells were measured to 1 µm by SEM using the software. In radulae, teeth were counted and total radula length measured to 0.1 mm.

Histology

Individual parts and organ systems − mainly reproductive organs and head-foot – of selected specimens were studied histologically. Specimens were embedded in paraffin using standard procedures. Female gonoducts containing embryonic shells had to be decalcified in 7% nitric acid (HNO3) first. Slide sections of 7–10 µm were stained with haematoxylin/eosine and embedded in Canada balsam.

Molecular genetics

DNA was purified from about 1–2 mm3 of foot tissue by CTAB extraction (Winnepenninckx, Backeljau & De Wachter, 1993). Polymerase chain reaction (PCR) was used to amplify a ∼890 bp region of the mitochondrial 16S ribosomal RNA gene using primers 16SF 5′−CCGCACTAGTGATAGCTAGTTTC and H3059 5′−CC GGTYTGAACTCAGATCATGT (Wilson et al., 2004). PCR was performed in 25 µL volumes containing 1X Taq buffer, 1.5 m m MgCl2, 200 µm each dNTP, 1–2.5 U Taq polymerase, c. 100 n m DNA and ddH2O up to volume on a Perkin Elmer GeneAmp 9600 thermocycler. After an initial denaturation step of 3 min at 94 °C, cycling conditions were 35 cycles of 1 min each at 94 °C,  50–55 °C  and  72 °C,  with  a  final  elongation step  of  5 min.  The  same  primers  were  used  in  PCR and sequencing. PCR products were purified with QiaQuick PCR purification kits (Qiagen). Both strands of both genes were cycle sequenced using ABI Prism BigDye terminator chemistry and visualized on an ABI Prism 377 automated DNA sequencer.

Sequences were aligned with Clustal X 1.8.1 for Windows (Thompson et al., 1997) using default settings. The resulting alignment was corrected manually. To determine the best substitution model for Bayesian inference analysis (see below), hierarchical likelihood ratio tests (Posada & Crandall, 2001) were carried out with MrModelTest 1.1 (Nylander, 2002) and PAUP*4.0b10a (Swofford, 1998).

Phylogenetic trees were reconstructed by maximum parsimony using the branch-and-bound algorithm as implemented in PAUP*, with gaps treated as fifth base. Support for nodes was estimated by bootstrap analysis (100 replicates). In addition, Bayesian inference (BI, Huelsenbeck et al., 2001) was employed to infer phylogeny by using MrBayes 2.01 (Huelsenbeck & Ronquist, 2001a, b). The MCMCMC algorithm was run with four independent chains for 250 000 generations, every 100th tree was sampled, and the 1500 first trees were discarded (burn-in).

Pseudopotamis supralirata (Smith, 1883) was chosen as outgroup to root the phylogeny, because Pseudopotamis has recently been identified as sistergroup of the Sulawesi pachychilids (see Glaubrecht & Rintelen, 2003).

Museum codens

  • ANSP Academy of Natural Sciences, Philadelphia

  • BMNH The Natural History Museum, London

  • MZB Museum Zoologi, Bogor, Indonesia

  • MNHN Museum d'Histoire Naturelle, Paris

  • NMB Naturhistorisches Museum, Basel

  • SMF Senckenberg-Museum, Frankfurt

  • ZMA Zoological Museum, Amsterdam

  • ZMB Museum für Naturkunde, Berlin (formerly Zoologisches Museum Berlin)

  • ZMZ Zoologisches Museum der Universität, Zürich

Anatomical abbreviations

  • aa anterior aorta

  • ag albumin gland

  • an anal papilla

  • ba buccal apparatus

  • bc buccal commissure

  • bg buccal ganglion

  • bp brood pouch

  • bv blood vessel

  • cc cerebral commissure

  • cg cerebral ganglion

  • cp capsule gland

  • cpc cerebro-pedal connective

  • cm columellar muscle

  • cn crescentic thickening

  • cr crescentic ridge

  • ct ctenidium

  • cv cardiovascular cavity

  • dg digestive gland

  • dgo digestive gland opening

  • em embryo

  • er esophageal roof

  • es oesophagus

  • ey eye

  • ft foot

  • gd gonoductal groove

  • gf glandular field

  • gg genital groove

  • gl glabella

  • go pallial gonoduct

  • gp glandular pad

  • gs gastric shield

  • ht heart

  • hy hypobranchial gland

  • il intestine loop

  • int intestine

  • kd kidney

  • lf longitudinal fold

  • ll lateral lamina

  • mc mantle cavity

  • me mantle edge

  • mf mantle fold

  • ml medial lamina

  • mo mouth

  • mr marginal fold

  • mu muscle

  • nd nidamental gland

  • nt nutritive tissue

  • od odontophore

  • op operculum

  • os osphradium

  • pg pedal ganglion

  • plg pleural ganglion

  • po pallial oviduct

  • ppc pleuro-pedal connective

  • pr posterior pouch (mantle cavity)

  • ra radula

  • rd radula sac

  • re rectum

  • rp nephridial pore

  • rs receptaculum seminis

  • sa sorting area

  • sb spermatophore bursa

  • sbg suboesophageal ganglion

  • sbc suboesophageal connective

  • sc statocyst

  • sg salivary gland

  • sn snout

  • spc supraoesophageal connective

  • srd subradular organ

  • ss style sac

  • st stomach

  • t1 major typhlosole

  • t2 minor typhlosole

  • te tentacle

  • tf transverse folds

  • ts testis

  • va vaginal opening

RESULTS

Systematic description of Tylomelania

The anatomical description presented here refers to all Sulawesi pachychilids as Tylomelania, i.e. including also species currently assigned to Brotia, in anticipation of the Discussion, where evidence from morphology and molecular genetics for a single pachychilid genus on Sulawesi is evaluated. Whenever, in contrast, reference is made to Tylomelania sensuSarasin & Sarasin (1897), this is clearly indicated. Table 1 lists all species-level taxa now included in Tylomelania and their former generic assignment. The description does not include a species-level treatment of taxa, as a revision is still in progress (T. van Rintelen, P. Bouchet & M. Glaubrecht, unpubl. data). All data presented are, unless explicitly stated, valid for all species of Sulawesi pachychilids examined so far and not influenced by taxonomic uncertainties at the species level. Specimens were assigned to species based on their shell and radula morphology by a comparison with the type specimens.

Caenogastropoda Cox, 1959 Cerithioidea Férussac, 1819 Pachychilidae Troschel, 1857 Tylomelania Sarasin & Sarasin, 1897

Tylomelania Sarasin & Sarasin, 1897: 317.

Type species: Tylomelania neritiformis Sarasin & Sarasin, 1897 (subsequent designation by Thiele, 1929)

Diagnosis

Shell globose to highly turreted, usually decollated, with 3–16 remaining whorls; colour yellowish or greenish-brown to black; smooth or with fine to strong axial and/or spiral ribs. Aperture oval, outer lip angulated or round, strong parietal callus in some species; anterior basis with slight extension. Embryonic shells with wrinkled sculpture on initial whorl; spiral striae, axial ribs and sometimes pronounced spiral grooves on subsequent whorls. Operculum oval to round and multispiral, with a central nucleus.

Head-foot grey or black, rarely yellow, dotted with white or yellow blotches in some species. Right mantle edge with broad upward-bound flap on inside. Radula usually typically pachychilid, very long and coiled at the posterior right of buccal mass; 2–5 denticles on marginals. Pallial gonoduct in both sexes open along its total length, with spermatophore bursa and receptaculum seminis in the female medial lamina. The species of the genus are ovoviviparous, with the pallial oviduct containing relatively few but very large juveniles with shells of multiple whorls, i.e. Tylomelania is a uterine brooder. Very large albumin gland at the posterior end of the pallial oviduct consists of glandular folds. Stomach typically pachychilid.

Endemic to Sulawesi, where no other pachychilids are known to occur.

Description

Shell (Figs 2, 3). Small to very large (adults 10–117 mm), yellowish-brown, brown or black. Turreted or elongately conic, in a few species globose, smooth or with fine to strong axial and/or spiral ribs. The first whorls usually eroded, often quite heavily, with 3–16 remaining whorls. Oval aperture, holostome, with a slight extension of the anterior outer lip in most species.

Figure 2

Tylomelania, shells (types) of all described lacustrine species. A−G, Lake Poso species. A, T. neritiformis (ST, NMB 1349a). B, T. connectens (HT, NMB 1335a). C, T. porcellanica (ST, NMB 1347°). D, T. carbo (ST, NMB 1343a). E, T. centaurus (HT, NMB 1339a). F, T. kuli (ST, NMB 1329a). G, T. toradjarum (ST, NMB 1328a). H−Z, Malili lake system species. H, T. insulaesacrae (ST, NMB 1342a). J, T. monacha (ST, NMB 1348a). K, T. zeamais (ST, NMB 1337a). L, T. gemmifera (ST, NMB 1344a). M, T. masapensis (ST, ZMA). N, T. tominangensis (ST, ZMA). O, T. lalemae (ST, ZMA). P, T. molesta (ST, NMB 1340a). Q, T. sarasinorum (ST, ZMA). R, T. abendanoni (ST, ZMA). S, T. towutica (ST, ZMA). T, T. bakara (HT, MZB Gst. 11.956). U, T. kruimeli (HT, MZB Gst. 11.959). V, T.palicolarum (ST, NMB 1331a). W, T. patriarchalis (ST, NMB 1330a). X, T. mahalonica (ST, ZMA). Y, T. mahalonensis (PLT, ZMA). Z, T. towutensis (PLT, NMB 4790b). Scale bar = 1 cm. HT – holotype, ST – syntype.

Figure 2

Tylomelania, shells (types) of all described lacustrine species. A−G, Lake Poso species. A, T. neritiformis (ST, NMB 1349a). B, T. connectens (HT, NMB 1335a). C, T. porcellanica (ST, NMB 1347°). D, T. carbo (ST, NMB 1343a). E, T. centaurus (HT, NMB 1339a). F, T. kuli (ST, NMB 1329a). G, T. toradjarum (ST, NMB 1328a). H−Z, Malili lake system species. H, T. insulaesacrae (ST, NMB 1342a). J, T. monacha (ST, NMB 1348a). K, T. zeamais (ST, NMB 1337a). L, T. gemmifera (ST, NMB 1344a). M, T. masapensis (ST, ZMA). N, T. tominangensis (ST, ZMA). O, T. lalemae (ST, ZMA). P, T. molesta (ST, NMB 1340a). Q, T. sarasinorum (ST, ZMA). R, T. abendanoni (ST, ZMA). S, T. towutica (ST, ZMA). T, T. bakara (HT, MZB Gst. 11.956). U, T. kruimeli (HT, MZB Gst. 11.959). V, T.palicolarum (ST, NMB 1331a). W, T. patriarchalis (ST, NMB 1330a). X, T. mahalonica (ST, ZMA). Y, T. mahalonensis (PLT, ZMA). Z, T. towutensis (PLT, NMB 4790b). Scale bar = 1 cm. HT – holotype, ST – syntype.

Figure 3

Tylomelania, shells (types) of all described riverine species. A, T. perfecta (ST, ZMZ 522348). B, T. robusta (ST, ZMA). C, T. celebicola (ST, NMB 1333a). D, T. scalariopsis (ST, NMB 1346a). E, T. carota (ST, NMB 1336a). F, T. wallacei (ST, BMNH 200220115). G, T. tomoriensis (ST, NMB 1345a). H, T. helmuti (HT, MZB Gst. 11.965). J, T. perconica (ST, NMB 3892a). Scale bar = 1 cm. HT – holotype, ST – syntype.

Figure 3

Tylomelania, shells (types) of all described riverine species. A, T. perfecta (ST, ZMZ 522348). B, T. robusta (ST, ZMA). C, T. celebicola (ST, NMB 1333a). D, T. scalariopsis (ST, NMB 1346a). E, T. carota (ST, NMB 1336a). F, T. wallacei (ST, BMNH 200220115). G, T. tomoriensis (ST, NMB 1345a). H, T. helmuti (HT, MZB Gst. 11.965). J, T. perconica (ST, NMB 3892a). Scale bar = 1 cm. HT – holotype, ST – syntype.

Operculum (op; Fig. 4B, C). Slightly oval to almost round, depending on the degree of increase in the last, smooth and hyaline whorl. Multispiral, with 5–11 whorls around a central nucleus. Ventral side has a large and round smooth attachment scar of the foot muscle.

Figure 4

Tylomelania, external morphology and operculum. A, T. sarasinorum, female (ZMB 190129). B, C, opercula, dorsal. B, T. perfecta (ZMB 190188). C, T. palicolarum (ZMA). D, E, T. perfecta (ZMB 190006). D, male, last coils. E, female. F−H, T. insulaesacrae, female (ZMB 190160). F, lateral right/dorsal. G, dorsal. H, ventral. J−L, T. palicolarum, female (ZMB 190153). J, lateral right/dorsal. K, lateral, left. L, ventral. Scale bar = 1 cm.

Figure 4

Tylomelania, external morphology and operculum. A, T. sarasinorum, female (ZMB 190129). B, C, opercula, dorsal. B, T. perfecta (ZMB 190188). C, T. palicolarum (ZMA). D, E, T. perfecta (ZMB 190006). D, male, last coils. E, female. F−H, T. insulaesacrae, female (ZMB 190160). F, lateral right/dorsal. G, dorsal. H, ventral. J−L, T. palicolarum, female (ZMB 190153). J, lateral right/dorsal. K, lateral, left. L, ventral. Scale bar = 1 cm.

External morphology (Fig. 4). The animal (head-foot) is grey or black, in some species dotted with irregular white or yellow blotches, rarely completely yellow. The broad snout (sn) with the mouth (mo) forms the anterior part of the head, the two tentacles (te) are situated at both sides of its base. The small black eyes (ey) are on lateral thickenings of the tentacle bases. In both sexes a genital groove (gg) runs along the foot along the right side of the head, it ends in a fold beneath the tentacle.

The posterior two-thirds of the head-foot are covered by the mantle, the mantle cavity occupies about one to 1.5 coils of the entire animal. The mantle edge (me) forms a distinct band and is pigmented, in some species its margin is serrated. Well developed papillae as in some thiarids have not been found. On the right side of the body an inward fold of the mantle forms a siphon-like structure (mf, Fig. 5A). This fold is positioned at the anterior end of the pallial gonoduct and might serve for releasing the gonadal products. The rectum (re) is visible through the mantle roof. In brooding females (Fig. 4A, E−L), the embryos (em) can be seen through the pallium as well, the full brood pouch (bp) displaces other organs of the pallial cavity or the immediate postpalliate area considerably, e.g. rectum and kidney (kd). The stomach (st) occupies most of the 3rd coil, comprising the intestine loop (il), style sac (ss) and sorting area (sa) with the crescentic ridge (cr). The digestive gland (dg) and the gonads (dorsal) fill the remaining 1–3 coils of the animal. In males (Fig. 4D), the light coloured and finely grained testis (ts, see also Fig. 6D) is about as large as the greenish-brown digestive gland and can easily be distinguished from it. In females, the minute ovary is branched and barely visible.

Figure 5

Tylomelania, pallial organs. A, B, female. A, pallial cavity, dorsal. T. perfecta (ZMB 190086). B, pallial oviduct, medial lamina. T. patriarchalis (ZMB 190050). C, male. Pallial gonoduct, interior. T. perfecta (ZMB 190086). Scale bar = 1 mm.

Figure 5

Tylomelania, pallial organs. A, B, female. A, pallial cavity, dorsal. T. perfecta (ZMB 190086). B, pallial oviduct, medial lamina. T. patriarchalis (ZMB 190050). C, male. Pallial gonoduct, interior. T. perfecta (ZMB 190086). Scale bar = 1 mm.

Figure 6

Tylomelania, histological sections of male reproductive organs. T. perfecta (ZMB 190086). Cross-sections, entire whorl, from anterior to posterior. A, last quarter of pallial gonoduct. B, last fifth of pallial gonoduct. C, immediately posterior of pallial gonoduct. D, testis and digestive gland, last whorl. Scale bar = 1 mm.

Figure 6

Tylomelania, histological sections of male reproductive organs. T. perfecta (ZMB 190086). Cross-sections, entire whorl, from anterior to posterior. A, last quarter of pallial gonoduct. B, last fifth of pallial gonoduct. C, immediately posterior of pallial gonoduct. D, testis and digestive gland, last whorl. Scale bar = 1 mm.

Pallial organs (Fig. 5A). The pallial organs are the osphradium (os), the ctenidium (gill, ct), the hypobranchial gland (hy), the rectum (see below, Alimentary system) and the pallial gonoduct (later in this section).

The white, straight and slender osphradium is embedded into the pallium; it is only about one third as long as the ctenidium. The ctenidium consists of about 80–120 triangular blades; their shape varies slightly along the gill. An efferent branchial vessel extends along the posterior part of the gill towards the heart. The voluminous rectum opens at the mantle edge with a free anal papilla (an). The long and narrow hypobranchial gland lies adjacent to the rectum, from shortly behind the anal papilla to the kidney.

The pallial gonoduct (go) is open, forming a deep and narrow groove between the medial lamina (ml) and the lateral lamina (ll, fused with the mantle or head-foot) in both sexes.

In males (Figs 5C, 6), the medial lamina is a simple fold without any pouches. The anterior parts of both laminae are glandular. The nidamental gland (nd) is formed by a longitudinally grooved glandular pad on the lateral lamina, and a glandular field (gf) on the medial lamina consisting of slightly inclined vertical ridges. The sperm duct enters the pallial gonoduct at the posterior end where medial and lateral lamina fuse.

In females (Figs 5A, B, 7, 8), the medial lamina has a slit-like vaginal opening (va) in the anterior third. A longitudinal fold (lf) within this opening leads to the formation of two separate tubes: the ventral receptaculum seminis (rs) and the dorsal spermatophore bursa (sb); (this interpretation of the dorsal tube is tentative, as no spermatophores have been found so far in Tylomelania, see Discussion). The gonoductal groove is divided by interlocked transverse folds (tf). This part of the pallial oviduct forms a uterine brood pouch in adult females, which seems to be formed by the capsule gland (cp). The embryonic shells in the brood pouch are each embedded in a lump of soft yellowish-white material between the transverse folds. The embryos cluster by size within the brood pouch; the largest one is found most anteriorly (Fig. 4E−L). The most posterior compartments of the embryo-containing part of the brood pouch hold either only minute embryos, starting with just 1−2 whorls, or eggs only; the surrounding tissue is considerably more compact than in the anterior part. A very large albumin gland (ag), occupying 10–20% of the whole length of the pallial oviduct, lies posterior to the embryo-containing section of the brood pouch. The albumin gland is build by lamellae originating from the lateral and medial lamella, respectively (Fig. 7G). The lower part forms a slit-like tube opening to the brood pouch, i.e. the gonoductal groove, if no embryos are present. The ovary duct enters at the end of the gonoductal groove into the albumin gland.

Figure 7

Tylomelania, histological sections of pallial oviduct. Cross-sections, from anterior to posterior. A−G, T. perfecta (ZMB 190006). A, section in anterior half of vaginal opening, foot and medial lamina only. B, Last third of vaginal opening. C, Brood pouch, first third. D, Brood pouch, centre. E, F, brood pouch, last half, section F c. 0.2 mm behind section E. G, Last fifth of pallial oviduct, at anterior edge of albumin gland. H, T. palicolarum (MNHN). Brood pouch, centre, with embryo. Scale bar = 1 mm.

Figure 7

Tylomelania, histological sections of pallial oviduct. Cross-sections, from anterior to posterior. A−G, T. perfecta (ZMB 190006). A, section in anterior half of vaginal opening, foot and medial lamina only. B, Last third of vaginal opening. C, Brood pouch, first third. D, Brood pouch, centre. E, F, brood pouch, last half, section F c. 0.2 mm behind section E. G, Last fifth of pallial oviduct, at anterior edge of albumin gland. H, T. palicolarum (MNHN). Brood pouch, centre, with embryo. Scale bar = 1 mm.

Figure 8

Tylomelania, schematic reconstruction of pallial oviduct. Pallial oviduct (bottom) and sketches of respective histological sections (top). Letters correspond to those in Fig. 7. The sections show only the medial lamina (A−F) and the albumin gland (G), not the whole gonoduct as in the preceding figure.

Figure 8

Tylomelania, schematic reconstruction of pallial oviduct. Pallial oviduct (bottom) and sketches of respective histological sections (top). Letters correspond to those in Fig. 7. The sections show only the medial lamina (A−F) and the albumin gland (G), not the whole gonoduct as in the preceding figure.

Embryonic shells(Fig. 9). The embryonic shells are relatively large; their size range is 2.8–17.5 mm (height of largest juvenile within brood pouch) with 2.8–7.5 whorls (N = 32 species). The initial whorl of the embryonic shells is wrinkled; in some specimens a transition of sculpture can be seen at about 1.3–1.5 whorls. At c. 2.5 whorls all embryonic shells start to show a pronounced axial sculpture, even in species which are smooth as subadults or adults. Two species-groups can be distinguished based on whether the axial ribs on their embryonic shells vanish again after around two whorls or not (Fig. 9A, B and D, E, respectively). Beside the axial ribs, faint to prominent spiral grooves may be present. The first whorls of the juveniles can be flat and rounded or high and steep; later whorls are always straight and angulated. The diameter of the first whorl (d) ranges between 225 and 375 µm (N = 39 species, including yet undescribed taxa).

Figure 9

Tylomelania, embryonic shells. A−C, T. perfecta (ZMB 190008). A, shell, lateral. B, shell, apical. C, protoconch, apical. D−F, T. matannensis (ZMB 190111). D, shell, lateral. E, shell, apical. F, protoconch, apical. Scale bars = 0.5 mm (A, B, D, E); = 30 µm (C, F).

Figure 9

Tylomelania, embryonic shells. A−C, T. perfecta (ZMB 190008). A, shell, lateral. B, shell, apical. C, protoconch, apical. D−F, T. matannensis (ZMB 190111). D, shell, lateral. E, shell, apical. F, protoconch, apical. Scale bars = 0.5 mm (A, B, D, E); = 30 µm (C, F).

Alimentary system . The buccal apparatus (Fig. 10A, B) immediately behind the mouth (mo) consists of the muscular (mu) buccal mass containing two jaws, the subradular organ (srd), and the odontophore (od) supporting the active part of the radula (ra), which is covered by the oesophageal roof (er). The oesophagus (es, Fig. 10C) exits the buccal mass posteriorly; its anterior section is covered by the large white to pinkish salivary glands (sg) in the area immediate posterior to the buccal mass. The long inactive part of the radula is coiled phonecord-like in the radula sac (rd) to the right and posterior of the buccal mass. The interior of the oesophagus is longitudinally folded.

Figure 10

Tylomelania, alimentary system. A, B, T. perfecta (ZMB 190086). A, buccal apparatus. B, buccal apparatus, detail of anterior section. C, stomach of T. patriarchalis (ZMB 190078). Scale bars = 1 mm.

Figure 10

Tylomelania, alimentary system. A, B, T. perfecta (ZMB 190086). A, buccal apparatus. B, buccal apparatus, detail of anterior section. C, stomach of T. patriarchalis (ZMB 190078). Scale bars = 1 mm.

The stomach of Tylomelania (Fig. 10C) has a narrow and elongate glandular pad (gp), a large and strongly concave cuticular gastric shield (gs) and a single crescentic ridge (cr). A single opening to the digestive gland (dgo) is present to the left of the glandular pad, anterior of the crescentic ridge. The crescentic thickenings (cn) of the sorting area (sa) are present in the stomach roof; the larger outer crescent has more pronounced septae than the inner one. A marginal fold (mr) marks the posterior border of the stomach. Intestine (int) and style sac (ss) communicate with each other, as major and minor typhlosole (t1, t2) are not fused. A crystalline style is present.

Radula(Fig. 11). The taenioglossate radula is generally very long and robust; the number of tooth rows per species ranges from 130 to 250 (mean species values). The maximum length found was 31.7 mm in T. patriarchalis (Sarasin & Sarasin, 1897). In smaller species, radula length can exceed shell height (e.g. 21.5 mm vs. c. 15 mm in T. carbo; Sarasin & Sarasin, 1897). A glabella (gl; Fig. 11A; cf. Troschel, 1857) is situated on the front of the central and the lateral teeth; in the former it resembles a ramp, in the latter a sail.

Figure 11

Tylomelania, radula. A, B, T. perfecta (ZMB 190080). A, segment, frontal. B, segment, apical (45 °). C, D, T. sarasinorum (ZMB 190123). C, segment, frontal. D, segment, apical (45°). E, F, T. carbo (ZMB 190200). E, segment, frontal. F, segment, apical (45 °). G, H, T. gemmifera (ZMB 190104). G, segment, frontal. H, segment, apical (45 °). Scale bars = 0.1 mm.

Figure 11

Tylomelania, radula. A, B, T. perfecta (ZMB 190080). A, segment, frontal. B, segment, apical (45 °). C, D, T. sarasinorum (ZMB 190123). C, segment, frontal. D, segment, apical (45°). E, F, T. carbo (ZMB 190200). E, segment, frontal. F, segment, apical (45 °). G, H, T. gemmifera (ZMB 190104). G, segment, frontal. H, segment, apical (45 °). Scale bars = 0.1 mm.

The rachidian is mostly squarish, with a highly variable set of denticles (cusps). Denticle arrangement ranges from a merely slightly enlarged central denticle, flanked by three minor denticles on each side (Fig. 11A, B), as in almost all other pachychilids from South-East Asia (Köhler & Glaubrecht, 2001), to an extremely prominent central denticle (Fig. 11C−F). In the latter case it can be either squarish or very pointed, with the minor denticles reduced in size and number or absent altogether. The upper and lower margins of the rachidian are concave and convex, respectively. In some species, two lateral ramp-like extensions running down from the lateral (minor) cusps converge towards the glabella. The lateral teeth are asymmetrical, the size and number of their denticles is correlated to those of the rachidian (cf. Fig. 11). The marginal teeth are usually to a varying extent hooked, the number of denticles varies from two to four on both interior and exterior marginals. The outer denticles are generally enlarged.

The radulae of two species are highly distinctive. T. gemmifera (Sarasin & Sarasin, 1897) (Fig. 11G, H; Lake Matano) and T. kuli (Sarasin & Sarasin, 1898) from Lake Poso have a very short radula (4–8 mm) with 20–30 tooth rows per mm, elongated laterals and unhooked marginals, giving it a thiarid-like appearance.

Renal system . The highly compartmentalized kidney lies posterior of the mantle cavity and anterior to the stomach. The mantle cavity has a pouch-like posterior extension (pr) with glandular ridges (Figs 6C, 7G), into which the kidney opens via a nephridial pore (rp, Fig. 6C).

Nervous system (Fig. 12). The epiathroid nerve system has a comparatively long cerebral commissure (cc) and, in contrast, rather short pleuro-pedal (ppc) and cerebro-pedal connectives (cpc). Cerebral (cg) and pleural ganglia (plg) are almost fused, the suboesophageal ganglion (sbg) is fused with the left pleural ganglion. The closely joined pedal ganglia (pg) are deeply embedded into the propodium muscle, covering the two lateral statocysts (sc), to which they are connected via short connectives. Seven major nerves branch off from each of the cerebral ganglia, innervating the snout (four nerves), tentacle, eye and, via a loop-shaped connective, the two buccal ganglia (bg). These are connected by a suboesophageal buccal commissure (bc). The anterior aorta (aa) passing through the nerve ring is very prominent and covers part of the supraoesophageal connective (spc) between the right pleural and the supraoesophageal ganglion.

Figure 12

Tylomelania, nervous system. Tylomelania perfecta (ZMB 190086). A, cerebral nerve ring, dorsal view. B, cerebral nerve ring, ventral view. C, buccal apparatus, left lateral view. D, buccal ganglia, from posterior. Scale bars = 1 mm.

Figure 12

Tylomelania, nervous system. Tylomelania perfecta (ZMB 190086). A, cerebral nerve ring, dorsal view. B, cerebral nerve ring, ventral view. C, buccal apparatus, left lateral view. D, buccal ganglia, from posterior. Scale bars = 1 mm.

Reproductive biology

All examined species of Tylomelania where sex determination was possible are gonochoristic (N = 30 species; N = 2300 individuals). The sex ratio, given as proportion of males among sexed individuals, ranges from 0.13 to 0.67 (cf. Table 2), averaging 0.42 overall, which is a highly significant deviation from a balanced sex ratio (χ2 = 53.261, P < 0.001). In several species or populations females are significantly larger than males (Table 3).

Table 2.

Sex ratio in selected Tylomelania species. Only species with respective sample sizes of N = 40 or more sexed individuals are listed. Population- level sex ratio ranges within one species have been given where at least two populations with N = 20 or more individuals have been sexed. Chi-square values from an asymptotic χ2 test. P = probability value; values in bold indicate significant deviations from a balanced sex ratio

Species sex ratio χ2 P N 
T. bakara 0.55  0.490  0.484  52 
T. gemmifera 0.26 31.030 <0.001 134 
T. helmuti 0.25 10.000 0.002  45 
T. insulaesacrae 0.39  2.373  0.123  52 
T. kruimeli 0.49  0.210  0.884  51 
T. mahalonensis 0.48  0.190  0.663  84 
T. masapensis 0.25 29.752 <0.001 121 
T. patriarchalis 0.48  0.934  0.334 243 
T. perfecta 0.47  1.344  0.246 373 
T. sarasinorum 0.48  0.476  0.490 369 
 population level 0.29/0.60   21/20 
T. towutensis 0.49  0.630  0.801 256 
 population level 0.42/0.67   132/36 
T. towutica 0.39 10.667 0.001 222 
T. zeamais 0.34 30.894 <0.001 315 
 population level 0.32/0.46  –  – 124/87 
Species sex ratio χ2 P N 
T. bakara 0.55  0.490  0.484  52 
T. gemmifera 0.26 31.030 <0.001 134 
T. helmuti 0.25 10.000 0.002  45 
T. insulaesacrae 0.39  2.373  0.123  52 
T. kruimeli 0.49  0.210  0.884  51 
T. mahalonensis 0.48  0.190  0.663  84 
T. masapensis 0.25 29.752 <0.001 121 
T. patriarchalis 0.48  0.934  0.334 243 
T. perfecta 0.47  1.344  0.246 373 
T. sarasinorum 0.48  0.476  0.490 369 
 population level 0.29/0.60   21/20 
T. towutensis 0.49  0.630  0.801 256 
 population level 0.42/0.67   132/36 
T. towutica 0.39 10.667 0.001 222 
T. zeamais 0.34 30.894 <0.001 315 
 population level 0.32/0.46  –  – 124/87 
Table 3.

Sexual size dimorphism in Tylomelania. In all listed species or populations males are smaller than females. Exclusion from this table does not necessarily mean that a Tylomelania species shows no sexual size dimorphism. Besides low sample sizes, taxonomic confusion may prohibit its statistical detection. h (3 last) = height of three last whorls, h = overall shell height, F = F-value from a one-factorial ANOVA, P = probability value; values in bold indicate highly significantly smaller males

Species Parameter F P N 
T. bakara h (3 last)  3.667  0.038 52 
T. kruimeli h (3 last)  8.890 0.002 60 
T. mahalonensis h (3 last)  7.954 0.007 85 
T. masapensis h (3 last)  8.137 <0.001 116 
T. sarasinorum 
ZMB 190123 10.956 <0.001 48 
ZMB 190134 16.895 <0.001 32 
Species Parameter F P N 
T. bakara h (3 last)  3.667  0.038 52 
T. kruimeli h (3 last)  8.890 0.002 60 
T. mahalonensis h (3 last)  7.954 0.007 85 
T. masapensis h (3 last)  8.137 <0.001 116 
T. sarasinorum 
ZMB 190123 10.956 <0.001 48 
ZMB 190134 16.895 <0.001 32 

On average, around 80% of females (N = 1216) contain embryos in their brood pouch, the number and size of which vary considerably within and between species (Table 4). As a rule, larger species have more and larger embryonic shells in their brood pouch. Consequently, the largest juvenile has been found in the largest species, T. patriarchalis, with an embryonic shell height of 17.5 mm in a female of c. 90 mm (h). In contrast, the largest number of juveniles has been found in a medium-sized riverine species, T. robusta, where a female of c. 40 mm (h) contained 39 juveniles with a maximum embryonic shell height of 4.8 mm.

Table 4.

Number and maximum size of embryonic shells in selected Tylomelania species. Only species with respective sample sizes of N = 10 or more investigated specimens are listed

Species Number of juveniles Maximum height of juveniles 
Range Mean SD N Range Mean SD N 
T. abendanoni 1–4 2.6 0.93 14 1.0–6.5  4.9 1.38 13 
T. bakara 5–17 9.0 3.29 21 6.5–10.0  7.7 0.91 21 
T. gemmifera 1–8 4.0 2.07 15 4.3–9.5  7.0 1.60 16 
T. helmuti 2–15 6.0 3.25 24 1.0–6.5  4.9 1.30 24 
T. insulaesacrae 2–15 5.7 2.99 15 2.8–4.5  3.6 0.54 11 
T. kruimeli 2–14 7.7 3.39 17 6.5–11.0  8.8 1.03 18 
T. mahalonensis 1–11 6.0 2.99 25 4.0–10.2  7.9 1.33 24 
T. masapensis 1–5 3.1 0.97 77 0.3–10.0  6.2 1.82 77 
T. patriarchalis 2–29 5.5 4.76 60 3.6–17.5  9.2 2.87 59 
T. perfecta 2–23 9.8 5.07 65 3.0–8.0  6.4 0.89 62 
T. sarasinorum 1–14 6.3 2.73 57 0.5–8.5  6.2 1.61 50 
T. solitaria 1–5 3.1 1.33 20 3.2–10.1  7.2 1.52 18 
T. towutensis 1–8 4.9 1.84 28 8.5–16.0 11.7 2.30 19 
T. towutica 1–7 4.9 1.79 15 0.5–9.3  6.9 2.18 15 
T. zeamais 1–10 4 : 3 1.87 54 1.0–7.6  5.1 1.03 54 
Species Number of juveniles Maximum height of juveniles 
Range Mean SD N Range Mean SD N 
T. abendanoni 1–4 2.6 0.93 14 1.0–6.5  4.9 1.38 13 
T. bakara 5–17 9.0 3.29 21 6.5–10.0  7.7 0.91 21 
T. gemmifera 1–8 4.0 2.07 15 4.3–9.5  7.0 1.60 16 
T. helmuti 2–15 6.0 3.25 24 1.0–6.5  4.9 1.30 24 
T. insulaesacrae 2–15 5.7 2.99 15 2.8–4.5  3.6 0.54 11 
T. kruimeli 2–14 7.7 3.39 17 6.5–11.0  8.8 1.03 18 
T. mahalonensis 1–11 6.0 2.99 25 4.0–10.2  7.9 1.33 24 
T. masapensis 1–5 3.1 0.97 77 0.3–10.0  6.2 1.82 77 
T. patriarchalis 2–29 5.5 4.76 60 3.6–17.5  9.2 2.87 59 
T. perfecta 2–23 9.8 5.07 65 3.0–8.0  6.4 0.89 62 
T. sarasinorum 1–14 6.3 2.73 57 0.5–8.5  6.2 1.61 50 
T. solitaria 1–5 3.1 1.33 20 3.2–10.1  7.2 1.52 18 
T. towutensis 1–8 4.9 1.84 28 8.5–16.0 11.7 2.30 19 
T. towutica 1–7 4.9 1.79 15 0.5–9.3  6.9 2.18 15 
T. zeamais 1–10 4 : 3 1.87 54 1.0–7.6  5.1 1.03 54 

A significant correlation between the size of females and the number of juveniles or the size of the largest embryonic shell contained in their brood pouch has been found in 6 and 9 populations, respectively, out of ten populations with sample sizes of N > 10 juvenile- carrying females tested (Table 5).

Table 5.

Correlation between female shell size and juvenile shell size and number; r = correlation coefficient (Pearson), P = probability value; values in bold indicate significant correlations

Species ZMB Locality Number of juveniles Largest juvenile (shell height) 
r P r P N 
T. bakara 190131 Lake Towuti 0.874 0.000 0.759 0.007 11 
T. masapensis 190208 Lake Masapi 0.564 0.000 0.421 0.000 76 
T. mahalonensis 190154 Lake Mahalona 0.639 0.004 0.599 0.009 18 
T. patriarchalis 190050 Lake Matano 0.101 0.494 0.758 0.000 47 
T. patriarchalis 190056 Lake Matano 0.285 0.268 0.556 0.031 15 
T. sarasinorum 190123 Lake Towuti 0.502 0.015 0.451 0.031 23 
T. sarasinorum 190134 Lake Towuti 0.691 0.027 0.644 0.044 10 
T. zeamais 190052 Lake Matano 0.113 0.533 0.818 0.000 29 
T. zeamais 190065 Lake Matano 0.529 0.061 0.814 0.000 14 
Species ZMB Locality Number of juveniles Largest juvenile (shell height) 
r P r P N 
T. bakara 190131 Lake Towuti 0.874 0.000 0.759 0.007 11 
T. masapensis 190208 Lake Masapi 0.564 0.000 0.421 0.000 76 
T. mahalonensis 190154 Lake Mahalona 0.639 0.004 0.599 0.009 18 
T. patriarchalis 190050 Lake Matano 0.101 0.494 0.758 0.000 47 
T. patriarchalis 190056 Lake Matano 0.285 0.268 0.556 0.031 15 
T. sarasinorum 190123 Lake Towuti 0.502 0.015 0.451 0.031 23 
T. sarasinorum 190134 Lake Towuti 0.691 0.027 0.644 0.044 10 
T. zeamais 190052 Lake Matano 0.113 0.533 0.818 0.000 29 
T. zeamais 190065 Lake Matano 0.529 0.061 0.814 0.000 14 

There is no evidence of seasonality in reproduction; the proportion of juvenile carrying females observed within and between the two field-work periods in August−September 1999 (end of dry season) and March 2000 (end of wet season) remained unchanged.

Habitat

The species of Tylomelania are either riverine (lotic), occurring in streams, small rivers and carstic resurgences, or lacustrine (lentic), with relatively many species in the two central lake systems of Sulawesi. No riverine species is found in the lakes and vice versa. Species in the relatively short stretches of river connecting the Malili lakes (Petea and Tominanga Rivers, Fig. 1) are considered lacustrine species.

The lotic species often occur in dense populations at waterfalls in limestone areas. They are confined to clear water biotopes and seem to be dependent on large fallen leaves rotting in the water. The snails are usually found attached to the underside of the leaves, where they most probably feed on algal films. They might also feed on the leaves themselves, although there is as yet no direct evidence of shredding.

The lacustrine Tylomelania species are specific to either soft or hard substrates, and found on sand and mud or rocks and sunken wood, respectively, down to a depth of at least 40 m. Population density is highest in shallow water (1–2 m), especially on rock. In some cases more than 100 animals per m2 were counted. Below −20 m snails become increasingly scarce. They appear to feed on algae, presumably diatoms, growing on the hard substrate or, to judge from sand in the rectum, on detritus when dwelling in the soft substrate.

Distribution

Endemic to Sulawesi with at least 34 species (Fig. 13, Table 1). Nine riverine species have been found in the streams and rivers of south, south-east and central Sulawesi. At least 25 species are endemic to the central lakes (Fig. 1): seven in Lake Poso (including Poso River) in central Sulawesi, and 18 in the Malili Lakes (Lakes Matano, Mahalona, Towuti, Lontoa and Masapi) in the easternmost part of south Sulawesi (including Tominanga River). Not found in north Sulawesi. Further details are given in Rintelen (2003).

Molecular phylogeny

The aligned 16S sequences have a length of 843 bp, with a very restricted number of short (1–5 bp) and largely unambiguous indels required by the inclusion of the outgroup sequence (Pseudopotamis supralirata). Of the 135 variable positions, 88 are parsimony informative. Uncorrected distances (p-distance) range from 0.1 to 17.3% across all sequences (mean 5.2%) and from 0.1 to 5.5% in the Sulawesi pachychilids (mean 3.8%).

The two phylogenetic analyses conducted employing maximum parsimony (MP) and Bayesian inference (BI) resulted in identical tree topologies (Fig. 14). The MP branch-and-bound search recovered six equally most parsimonious trees with a length of 306 steps resulting in a fairly well resolved strict consensus cladogram (Fig. 14A). The BI phylogram differs only in showing a basal polytomy within the Sulawesi pachychilids (Fig. 14B).

Figure 14

Tylomelania, molecular phylogeny. Based on 890 bp of the mitochondrial 16SrRNA gene. Grey boxes and taxa names in bold type indicate the three species of Tylomelania sensuSarasin & Sarasin (1897, 1898). A, strict consensus cladogram of six equally most parsimonious trees. Numbers on branches indicate bootstrap support. B, Bayesian inference phylogram. Numbers on branches are posterior probabilities.

Figure 14

Tylomelania, molecular phylogeny. Based on 890 bp of the mitochondrial 16SrRNA gene. Grey boxes and taxa names in bold type indicate the three species of Tylomelania sensuSarasin & Sarasin (1897, 1898). A, strict consensus cladogram of six equally most parsimonious trees. Numbers on branches indicate bootstrap support. B, Bayesian inference phylogram. Numbers on branches are posterior probabilities.

While the Sulawesi pachychilids form a monophyletic clade, Tylomelania sensuSarasin & Sarasin (1897), comprising the species T. carbo, T. connectens and T. neritiformis, is shown to be para- or even polyphyletic. Its constituent species are found twice in different terminal positions nested within one of the four major subdivisions.

DISCUSSION

Taxonomy of the Sulawesi pachychilids

The systematic position of the lake gastropods and their fluviatile allies remained enigmatic for a long time. Based on the multispiral operculum with a central nucleus, Sarasin & Sarasin (1898) described all new species from the central lakes, including Tylomelania, as ‘Palaeomelanien’ in order to distinguish them from all other ‘melaniids’ found on Sulawesi. The latter, termed ‘Neomelanien’, have a paucispiral operculum with an excentric nucleus. The Sarasins regarded both as informal subgroups of ‘Melaniidae’ (see below and Glaubrecht, 1996, 1999 for details of the recent classification of this group). Seven riverine species, including the widespread ‘Melania’ perfectaMousson, 1849 were also placed in the ‘Palaeomelanien’ based on their operculum. Adding the lake species later described by Kruimel (1913), this group comprises a total of 31 (sub) species known at that time (Table 1). Thus Sarasin & Sarasin (1898: 8) explicitly refrained from the assignment of the ‘Melania’ species among their ‘Palaeomelanien’ to one of the already named melaniid genera set up for species with a round and multispiral operculum (e.g. from Asia, Sulcospira Troschel, 1857, Brotia H. Adams, 1866 and PseudopotamisMartens, 1894) as they believed a decision for one of these genera was not possible based on the data available at the time. They pointed to the example of South-east Asian Brotia, originally erected solely based on a multispiral operculum, which is not a character exclusive to it. The Sarasins noted, however, that the radula of the ‘Palaeomelanien’ was of the same type as the one described by Troschel (1857) for his ‘Pachychili’, one of four groups he distinguished among the ‘Melaniidae’, comprising PachychilusLea & Lea, 1850 from South America and Sulcospira.

Fischer & Crosse (1892), taking Troschel's concept further, divided the ‘Melaniidae’ into six groups, one of which was the Pachychilinae. Thiele (1921) replaced the name Pachychilinae, which he erroneously considered to be invalid, with Melanatriinae. These implicitly comprised the ‘Palaeomelanien’ of Sulawesi. Subsequently, Thiele (1925, 1928,,1929) delimited the Melanatriinae based on radula and operculum characters. Köhler & Glaubrecht (2001) replaced Melanatriinae with the original Pachychilidae. On the generic level, Thiele (1928) used the similarities of radula and operculum to assign the Sulawesi species described as ‘Melania’ to Brotia. He regarded Tylomelania as a subgenus of Sulcospira, again using radula characters, and the parietal callus of the shell aperture (Thiele, 1929). This in many respects ‘modern’ system, however, was largely ignored by later authors until recently, and all ‘melanians’ were subsumed under the Thiaridae (e.g. Morrison, 1954).

Based on morphology, the Pachychilidae were hypothesized to form a monophyletic group outside the Thiaridae sensu stricto (Glaubrecht, 1996, 1999; Köhler, Rintelen & Glaubrecht, 2000; Köhler & Glaubrecht, 2003).  Results  of  a  molecular  phylogeny of Cerithioidea (Lydeard et al., 2002) support this hypothesis, as does a cladistic analysis of morphological data (Glaubrecht, unpubl. data).

For a long time, it has been assumed that the species assigned here to Tylomelania have their closest relatives among the other South-east Asian pachychilids (see above and Introduction). This assumption was only made implicitly, however, by the generic assignation of the two groups then recognized on Sulawesi − i.e. Tylomelania sensuSarasin & Sarasin (1897, 1898) and ‘Melania’− to the two widespread South-east Asian genera Brotia and Sulcospira by Thiele (1928, 1929). For most of the 20th century this systematic scheme was not challenged; in the worst case it was simply ignored and the invalid genus ‘Melania’ invoked again (Marwoto, 1997).

The assignment of the Sulawesi pachychilids to Brotia and Sulcospira was recently challenged on morphological grounds and the existence of three independent pachychilid lineages in South-east Asia recognized (Köhler & Glaubrecht, 2001). Brotia comprises species from the Asian mainland and Sumatra only, while Sulcospira is confined to Java (Köhler, 2003). In contrast, the Sulawesi pachychilids are an independent lineage, a monophyletic group, based on both morphological (Rintelen & Glaubrecht, 1999; this study) and molecular data (Glaubrecht & Rintelen, 2003; this study). Consequently, we suggest extending the concept of Tylomelania, originally erected for four taxa endemic to Lake Poso (Sarasin & Sarasin, 1897, 1898), to comprise all pachychilid species on Sulawesi, thus including taxa formerly assigned to Brotia. This is based on the following evidence.

Morphology. The description of Tylomelania by Sarasin & Sarasin (1897) is based on two morphological characters: (1) a globose shell with a parietal callus and (2) a radula with a single vastly enlarged central denticle on the rachidian and the lateral teeth (Sarasin & Sarasin, 1898). The shell − in gastropods in general and in Sulawesi pachychilids in particular − is highly variable. Even within Tylomelania (sensu Sarasin & Sarasin), there is considerable variation. For example, T. carbo is quite distinct from the other three species. A comprehensive study of radula morphology in almost all species of Sulawesi pachychilids, which will be published in the course of a species revision (Rintelen, Bouchet & Glaubrecht, unpubl. data), has revealed that the ‘unique’ radula of Tylomelania sensu Sarasin & Sarasin is well within the interspecific variation range of the whole Sulawesi species flock (Fig. 11). Thus, the morphological data render the separation of a subset of species in a different genus based on these two characters a rather arbitrary decision.

Soft-part anatomy . All pachychilid species of Sulawesi have a very similar soft-part anatomy as described above (Figs 412), with virtually no specific differences in qualitative traits at the generic level. Thus, anatomical characters do not support a subdivision of the Sulawesi pachychilids.

Molecular genetic data. Although suggesting the existence of several subdivisions within the Sulawesi pachychilids (Fig. 14), these do not support a clade corresponding to Tylomelania sensuSarasin & Sarasin (1897, 1898), which apparently is para- or polyphyletic.

In summary, the Sulawesi pachychilids cannot be morphologically subdivided without making a highly arbitrary decision, disregarding the criterion of monophyly, if Tylomelania is restricted to the original set of species. Alternatively, the erection of several new genera based on the molecular phylogeny would hardly be supported by morphological characters and would certainly not include Tylomelania in its original sense. Consequently, all species from Sulawesi formerly assigned to Brotia are here transferred to Tylomelania, which is the oldest available generic name exclusive to these endemic pachychilids (Table 1). Thus, as one of three pachychilid lineages in South-east Asia (Köhler & Glaubrecht, 2001, 2003; see below), Tylomelania is the only pachychilid genus in Sulawesi.

Based on molecular data and preliminary morphological findings, the pachychilid genus PseudopotamisMartens, 1894 with only two species in the North Australian Torres Strait Islands has recently been proposed as the sister group of Tylomelania (see below; Glaubrecht & Rintelen, 2003).

Comparative and phylogenetic aspects of the anatomy of Tylomelania

General anatomy

The anatomy of Tylomelania as described above is generally similar to that of all other pachychilid genera (see e.g. descriptions of African Potadoma Swainson, 1840 (Binder, 1959), Malagassy Melanatria Bowdich, 1822 (Starmühlner, 1969), South American Pachychilus (Simone, 2001) and the South-east Asian Brotia and Jagora (Köhler & Glaubrecht, 2001, 2003). The very different reproductive system is the most obvious exception and is discussed in detail below.

Of the remaining characters none are possible autapomorphies of Tylomelania, whose morphology is dominated by those features that are likely autapomorphies of the Pachychilidae, i.e. symplesiomorphies at the level discussed here. These include the more-or-less round, multispiral operculum with a central nucleus, the mantle flap on the right side of the head-foot, the pouch-like posterior extension of the mantle cavity containing the nephropore, the rather long coiled radula, and the stomach as described above. The phylogenetic value of characters in the nervous system of Tylomelania is difficult to assess, as a comparative study of pachychilid nervous systems is still lacking. The only nervous system of a South-east Asian pachychilid depicted so far, Brotia pageli (Thiele, 1908), by Köhler & Glaubrecht (2001) does not seem to differ from that of Tylomelania. This finding needs to be verified, as do the small differences between the nervous systems of Tylomelania and Melanatria (Starmühlner, 1969) in the connection of the statocysts and nerves branching off from the cerebral ganglia.

Apart from the reproductive system, the only synapomorphy of Tylomelania and its sister taxon Pseudopotamis found so far is a radula with primarily three marginal denticles (Glaubrecht & Rintelen, 2003; this study).

Reproductive anatomy and biology

In contrast to all characters mentioned so far, the female reproductive system of Tylomelania is fundamentally different from that of all other pachychilids except Pseudopotamis. The sole non-derived character of this organ complex is the open gonoduct (in both sexes), a condition which is, however, considered apomorphic for Cerithioidea in general (Glaubrecht, 1996).

The pallial oviduct of Tylomelania is modified to a uterine brood pouch, a probable synapomorphy with Pseudopotamis, even though detailed studies of the latter are still lacking (see below; Glaubrecht & Rintelen, 2003). The modified pallial oviduct of Tylomelania is indeed a complex of several associated derived characters; these include the apparent modification of the capsule gland into a compartmentalized brood pouch, the large albumin gland, and the medial lamina with the specific arrangement of receptaculum seminis and spermatophore bursa (cf. Figs 5, 7, 8). The existence of a compartmentalized uterine brood pouch has been confirmed for Pseudopotamis, while the arrangement of a spermatophore bursa and receptaculum seminis in the medial lamina as well as an albumin gland might differ from that of Tylomelania (Glaubrecht & Rintelen, 2003; unpubl. data).

The embryonic shells of Tylomelania are unique among pachychilids (including Pseudopotamis) in having axial ribs irrespective of adult shell sculpture (Fig. 9). Their apical whorl has a slightly wrinkled surface typical of several viviparous cerithioideans (see e.g. Glaubrecht, 1996; Köhler & Glaubrecht, 2001). However, in contrast to other viviparous pachychilids such as Brotia and Jagora, there is no evidence for a retarded formation of the apical shell due to consumption of yolk through the apex, since even the smallest observable embryonic shells in the brood pouch have a calcified apex.

As outlined above, the embryonic shells in the brood pouch of Tylomelania are embedded in nutritive material contained within the egg capsule. The nutritive function of the tissue can be assessed by comparing the amount present around the largest, most anterior embryos (where it is almost gone) and the smaller, more posterior shells, which are still deeply embedded in it. Thus, there is no direct transfer of nutrients to the embryos during their growth period, e.g. through a histotrophe (see below). Instead, these findings suggest that the huge, conspicuous albumin gland is responsible for providing the nutrients for development within the egg capsules.

Tylomelania has the largest embryos (size at birth) among freshwater gastropods (cf. Dillon, 2000: 139) and possibly all viviparous gastropods. Moreover, the size range (largest embryo at birth) within Tylomelania exceeds at least that of all other live-bearing cerithioideans as well (Table 6).

Table 6.

Diversity of breeding structures and strategies in Cerithioidea. Only species releasing shelled crawling juveniles have been listed. Asterisks indicate the poecilogonic species Planaxis sulcatus, where numbers separated by a semicolon represent two different reproductive strategies. For the embryonic parameters the range could not always be given due to lack of data. Abbreviations: SHE, shell height of largest embryo (mm); AAS, average adult shell height

Taxon Brooding type (brood pouch) Nourishment No. juveniles SHE AAS Source 
Cerithioidea 
Marine 
Planaxidae 
  Planaxis sulcatus subhaemocoel nurse eggs 3–4; c. 600* 3.25; 0.5* 19.5 Glaubrecht (1996); unpubl. data 
Freshwater 
Pachychilidae 
  Adamietta hainanensis subhaemocoel lecithotroph 66–248 1.0–1.5 34.2 Köhler & Glaubrecht (2001); Köhler (2003
  Brotia pagodula subhaemocoel lecithotroph 1–31 3.0–5.6 27.4 Köhler & Glaubrecht (2001); Köhler (2003
  Jagora asperata pallial lecithotroph 275 1.7–2.0 48.4 Köhler & Glaubrecht (2003
Tylomelania insulaesacrae uterine lecithotroph 2–15 2.8–4.5 14.4 this study 
  Tylomelania patriarchalis uterine lecithotroph 2–29 3.6–17.5 60.8 this study 
Paludomidae 
  Lavigeria nassa uterine lecithotroph 43 1.5 14.4 Glaubrecht, unpubl. data 
  Potadomoides pelseneeri uterine lecithotroph 195 0.4 10.0 Glaubrecht & Strong, unpubl. data 
  Tanganyicia rufofilosa mesopodial lecithotroph 68 1.0 16.0 Strong & Glaubrecht (2002
  Tiphobia horei uterine lecithotroph 36 3.2 41.5 Strong & Glaubrecht, unpubl. data 
Pleuroceridae 
  Semisulcospira libertina uterine lecithotroph 6–1016 2.4 31.0 Davis (1972); Takami (1991) 
Thiaridae 
  Hemisinus cf. tuberculatus subhaemocoel matrotroph 14 2.0–3.0 37.5 Glaubrecht (1996
  Melanoides tuberculata subhaemocoel matrotroph 15–265 4.3 30.0 Berry & Kadri (1974); Glaubrecht (1996
  Tarebia granifera subhaemocoel matrotroph 25–74 3.0 25.0 Glaubrecht (1996
  Thiara scabra subhaemocoel matrotroph 75–110 3.4 19.0 Riech (1937); Glaubrecht (1996); unpubl. data 
Selected other non-pulmonate 
gastropods 
Littorinidae 
Littorina saxatilis uterine lecithotroph 10–900 0.4–1.4 20.0 Reid (1996) 
Viviparidae 
  Viviparus viviparus uterine lecithotroph 6–96 4.0 32.0 Frömming (1956); Fretter & Graham (1994
Taxon Brooding type (brood pouch) Nourishment No. juveniles SHE AAS Source 
Cerithioidea 
Marine 
Planaxidae 
  Planaxis sulcatus subhaemocoel nurse eggs 3–4; c. 600* 3.25; 0.5* 19.5 Glaubrecht (1996); unpubl. data 
Freshwater 
Pachychilidae 
  Adamietta hainanensis subhaemocoel lecithotroph 66–248 1.0–1.5 34.2 Köhler & Glaubrecht (2001); Köhler (2003
  Brotia pagodula subhaemocoel lecithotroph 1–31 3.0–5.6 27.4 Köhler & Glaubrecht (2001); Köhler (2003
  Jagora asperata pallial lecithotroph 275 1.7–2.0 48.4 Köhler & Glaubrecht (2003
Tylomelania insulaesacrae uterine lecithotroph 2–15 2.8–4.5 14.4 this study 
  Tylomelania patriarchalis uterine lecithotroph 2–29 3.6–17.5 60.8 this study 
Paludomidae 
  Lavigeria nassa uterine lecithotroph 43 1.5 14.4 Glaubrecht, unpubl. data 
  Potadomoides pelseneeri uterine lecithotroph 195 0.4 10.0 Glaubrecht & Strong, unpubl. data 
  Tanganyicia rufofilosa mesopodial lecithotroph 68 1.0 16.0 Strong & Glaubrecht (2002
  Tiphobia horei uterine lecithotroph 36 3.2 41.5 Strong & Glaubrecht, unpubl. data 
Pleuroceridae 
  Semisulcospira libertina uterine lecithotroph 6–1016 2.4 31.0 Davis (1972); Takami (1991) 
Thiaridae 
  Hemisinus cf. tuberculatus subhaemocoel matrotroph 14 2.0–3.0 37.5 Glaubrecht (1996
  Melanoides tuberculata subhaemocoel matrotroph 15–265 4.3 30.0 Berry & Kadri (1974); Glaubrecht (1996
  Tarebia granifera subhaemocoel matrotroph 25–74 3.0 25.0 Glaubrecht (1996
  Thiara scabra subhaemocoel matrotroph 75–110 3.4 19.0 Riech (1937); Glaubrecht (1996); unpubl. data 
Selected other non-pulmonate 
gastropods 
Littorinidae 
Littorina saxatilis uterine lecithotroph 10–900 0.4–1.4 20.0 Reid (1996) 
Viviparidae 
  Viviparus viviparus uterine lecithotroph 6–96 4.0 32.0 Frömming (1956); Fretter & Graham (1994

Interestingly, no evidence for seasonality in reproduction was observed. Tylomelania species can obviously be long-lived, as suggested by the very large and strong shells which require some time to build, especially in the oligotrophic lakes of Sulawesi. The ‘layout’ of the brood pouch suggests continuous reproduction, where only the largest embryos are released. If this hypothesis could be confirmed, Tylomelania would have a life-cycle which is very different from that of other South-east Asian pachychilids, where evidence for reproductive cohorts was found (Dudgeon, 1982, 1989; Köhler & Glaubrecht, 2003). While ignored in general treatments of freshwater gastropod life-cycles (Calow, 1978; Dillon, 2000), some thiarids seem to reproduce continuously as well (see review in Glaubrecht, 1996). However, data on the life cycle of Tylomelania species are still fragmentary at best. Nothing is yet known about growth rates, age at maturity and life span. Equally lacking are data on the paternity of the embryos, i.e. the question of whether all offspring in a brood pouch have been sired by one or several males. The failure to find spermatophores in Tylomelania (although more than 50 males have been dissected) or in any other pachychilid, is conspicuous. In the absence of a penis their presence is, however, highly likely, which is why pouches in the medial lamina of the discussed taxa are nevertheless regarded as spermatophore bursae. Among freshwater Cerithioidea an almost bizarre variety of forms has been described in the Paludomidae of Lake Tanganyika (Glaubrecht & Strong, 2004).

Tylomelania is gonochoristic, as are all other South-east Asian pachychilids, in contrast to earlier suggestions by several authors (see review in Köhler & Glaubrecht, 2001). While the overall sex ratio deviates significantly from 0.5 (a completely balanced ratio) in favour of females, a wide range of ratios was observed among species and populations (Table 2). A discussion and comparison of these findings is difficult, as non-random sampling in respect to sex cannot be excluded (cf. criteria given by Dillon, 2000: 109). Females are often larger than males, and therefore probably more likely to be collected and examined. This suspicion is supported by the sex ratios of those species where a large number of individuals was sexed (e.g. T. patriarchalis, T. perfecta, T. sarasinorum; Table 2), which are usually not significantly different from 0.5. In just one species (T. masapensis) with a highly significantly skewed sex ratio, this is most probably not an artefact, as the sample was both randomly collected and examined. If present, the reason for such skewed sex ratios remains speculative. In other freshwater gastropods a positive correlation between the proportion of males and the degree of parasite infestation has been observed (Johnson, 1994; Lively & Johnson, 1994; Dybdahl & Lively, 1996). A similar causation might be suspected for Tylomelania as well, as some populations were found to be heavily parasitized.

In contrast to the uterine brooder Tylomelania, Brotia and the closely related Adamietta Brandt, 1974 have a subhaemocoelic brood pouch in the head-foot which is very similar in both taxa, while their pallial gonoducts are distinct (Köhler & Glaubrecht, 2001; Köhler, 2003). Also in contrast to Tylomelania, all of these taxa have clearly detectable capsule glands in the dorsal area of the pallial oviduct where lateral and medial lamina fuse, and an albumin gland has not been found. The position of the receptaculum seminis and spermatophore bursa in the medial lamina of Brotia pageli is analogous to that seen in Tylomelania, while in the other genera it is not even remotely similar (Köhler & Glaubrecht, 2001). The Philippine endemic Jagora has a fundamentally different reproductive system again (Köhler & Glaubrecht, 2003). Its pallial oviduct has a unique arrangement of spermatophore bursa and receptaculum seminis. The position of the capsule gland is the same as for the other South-east Asian taxa; in addition, an albumin gland is present at the posterior end of the gonoduct. Furthermore, it retains egg capsules in the mantle cavity and provides a rather simple form of brood protection.

The variety of brooding strategies among pachychilids is extraordinary. Three different morphological solutions have evolved, with two different life-history strategies (see above). Tylomelania and Pseudopotamis clearly follow a ‘k-strategy’ (MacArthur & Wilson, 1967) and produce comparatively few large embryos with a lot of reproductive effort devoted to each, while the other South-east Asian pachychilids generally have many small juveniles, up to 800 in some species (‘r-strategy’; see Discussion in Köhler & Glaubrecht, 2001). It should be noted that this comparison of life-history traits holds only true within the pachychilids, while the concept is relative (see Glaubrecht, 1996). In comparison to some thiarids with a large number of veligers in the brood pouch (e.g. Stenomelania), all live-bearing pachychilids are ‘k-strategists’ (Köhler & Glaubrecht, 2003).

Based on these findings, the homology of the reproductive systems of Tylomelania and the South-east Asian pachychilid genera must be reassessed. The general features of the gonoduct are without doubt homologous (open in both sexes, possession of receptaculum seminis and spermatophore bursa in the medial lamina), and the respective characters are symplesiomorphic within the Pachychilidae. The position of the spermatophore bursa and receptaculum seminis (including their openings within the medial lamina) and of the gonads and the digestive gland in the last few body whorls is probably autapomorphic, at least for each viviparous pachychilid lineage (Köhler, 2003). However, the lack of respective data for some oviparous taxa (Melanatria, Potadoma) impedes a more decisive statement. In contrast to the gonoducts, the brooding structures (as such) of the viviparous pachychilids are not considered homologous here, in agreement with findings by Rintelen & Glaubrecht (1999) and Köhler & Glaubrecht (2001). From a phylogenetic point of view the brooding structures described above each represent autapomorphies of the respective taxa. However, the reproductive system, including the brood pouch of Pseudopotamis, very likely represents a synapomorphy of that taxon and Tylomelania, and is indicative of a sister-group relationship of both taxa (Glaubrecht & Rintelen, 2003).

Viviparity or ovoviviparity?

Viviparity has hitherto been used as a synonym for giving birth to living (shelled) juveniles. However, the variety of pachychilid breeding structures requires some clarification of the term, as there are clear differences between the protection of egg capsules in the mantle cavity (as in Jagora) and the maturation of juveniles in very large amounts of albumen within the pallial oviduct (as in Tylomelania). A common criterion of viviparity in a strict sense is the direct nutrient transfer from the mother animal to the juveniles via a histotrophe (e.g. Fretter & Graham, 1994: 663), which is also referred to as matrotrophic viviparity (see, e.g. Meyer & Lydeard, 1993; Glaubrecht, 1996; Korniushin & Glaubrecht, 2003). All ‘viviparous’ taxa in which such direct transfer of nutrients to the growing embryo is lacking are consequently ovoviviparous (lecithotrophic viviparity).

Accordingly, all pachychilid breeding strategies discussed here should be regarded as ovoviviparous (Table 6). No secretive nutritive tissue could be observed, and the nutrients (albumen) are provided to the embryo within the egg capsule in all cases, which are retained within the body (Köhler & Glaubrecht, 2001, 2003; this study). In contrast to earlier assumptions (see review in Glaubrecht, 1996), true viviparity among Cerithioidea is only found in the Thiaridae (Table 6; Glaubrecht, 1996, 1999; Köhler & Glaubrecht, 2001). Even within the wider limits of the Caenogastropoda only one other (uncertain) case of true viviparity can be found (in Janthina; Fretter & Graham, 1994).

As shown by the variety of brooding structures in Cerithioidea (Table 6), the term ‘ovoviviparous’ should not be used to indiscriminately lump together the wide range of complexity and maternal investment in the taxa thus subsumed. While ovoviviparous taxa with more sophisticated reproductive modes (e.g. Tylomelania or Brotia) may exceed the viviparous Thiaridae in maternal investment and have structures that are almost as complex, the simple mode found in Jagora barely surpasses the stage of simple brood care. Additional terminological distinctions might thus help to discriminate between such extremes for certain purposes such as morphological comparisons. However, for questions of more general biological interest, such as the effect of brooding on speciation (see next section) the distinction between matrotrophy and lecithotrophy has been neglected, as generally the stress is on contrasting the release of living young (viviparity in a broad sense) and egg-laying (oviparity).

Viviparity and gastropod radiations in ancient lakes

Viviparity (including ovoviviparity as differentiated herein) has frequently been discussed as an important factor in the diversification of gastropod species flocks in ancient lakes (reviews in Boss, 1978; Michel, 1994; Glaubrecht, 1996). It has been claimed for one genus (Lavigeria) among the 36 gastropod species constituting the Lake Tanganyika flock that slowly dispersing brooding clades show a higher degree of interpopulation differentiation, even without dispersal barriers, than non-brooding ones (Cohen & Johnston, 1987; Johnston & Cohen, 1987). Moreover, the ability to speciate allopatrically seems to be linked with the potential to diversify morphologically as well, as evidenced by a higher number of sympatric species in Lavigeria (Michel, 2000; West et al., 2003).

While Tylomelania cannot be employed to test the effects of brooding on speciation and diversification, as all species share the same reproductive strategy, instructive comparisons could potentially be made with other brooding cerithioidean taxa. Around the edges of all Sulawesi lakes, thiarids − mainly Tarebia granifera (Lamarck, 1822) and Melanoides tuberculata (Müller, 1774) − occur in streams and rivers. They are often syntopic with the widespread fluviatile Tylomelania perfecta (Table 1). Only T. granifera is found in a few localities in the lake itself, restricted to extremely shallow and not very clear water. Both thiarid species are euviviparous (Table 5; Glaubrecht, 1996, 1999) and parthenogenetic. This finding indicates that rather than the ability to brood, sexual reproduction or competitive exclusion might be important factors in intralacustrine radiations. Similar observations can be made in Lake Tanganyika, which contains only gonochoristic species, while the so-called radiation of Melanoides in Lake Malawi might provide a possible exception (Brown, 1994).

Biogeography of the Sulawesi pachychilids

As outlined above, the North Australian genus Pseudopotamis, an endemic with two species on the Torres Strait Islands, has been proposed as sister group of Tylomelania based on morphological and molecular data (Glaubrecht & Rintelen, 2003). The results of phylogenetic analyses of larger molecular data sets (Köhler, 2003; Rintelen, 2003) provide almost unanimous support for this hypothesis, confirming the adelphotaxon status of both genera. However, the biogeographical implications of this finding present a potentially significant problem. Sulawesi and the Torres Strait Islands are separated by a 2000 km stretch of sea. The present-day pattern must therefore be explained by invoking either a vicariance or a dispersal hypothesis.

Vicariance appears to be the obvious choice, as suggested both by the molecular pattern and the presumably extremely low cross-ocean dispersal ability of these strictly freshwater-dwelling ovoviviparous taxa (Glaubrecht & Rintelen, 2003). However, in order to assess the support for a vicariance hypothesis which might explain a Sulawesi–Australia distribution pattern, the geological background is essential. Sulawesi is situated in one of the geologically most complex regions of the world. While it is very difficult for the non-geologist to fully understand the finer points of South-east Asian geological history, a basic comprehension of the geological constraints imposed on the area is crucial to a meaningful interpretation of biogeographical patterns.

South-east Asia consists of Gondwanan fragments of various ages, which collided with Eurasia in three sequential events during the Palaeozoic and Mesozoic, from the Devonian (350 Mya) to the end of the Jurassic (140 Mya) (Metcalfe, 1996,1998, 2001). During the Cretaceous, India and Australia separated from Gondwana and moved northwards. The collision of India with Asia c. 50 Mya caused major deformations in South-east Asia; these continue to the present day. Australia's collision with the Sundaland margin caused rapid changes in topography in eastern Indonesia and the adjacent West Pacific, especially during the last 25 Myr (Hall, 1996,,1998, 2001). Thus, geological evolution during the Cenozoic (65 Mya to present) is of great relevance to biogeographical questions, as the major tectonic events determining the present configuration of land and sea in the area occurred during that period (Hall, 2001).

Sulawesi consists of four tectonic provinces with continental fragments – also referred to as microplates or terranes – at their core (Hall, 1996,,1998; Moss & Wilson, 1998; Wilson & Moss, 1999). These provinces are, from west to east: (1) the West Sulawesi Plutonic-Volcanic Arc, comprising the south arm, the western part of central Sulawesi and the neck of the north arm; (2) the Central Sulawesi Metamorphic Belt, comprising the southern part of the south-east arm and the middle of central Sulawesi; (3) the East Sulawesi Ophiolite Belt, comprising the east arm and the north-eastern part of the south-east arm, and (4) microcontinental fragments forming the smaller adjacent islands of Banggai and Sula, and the Tukang Besi group. In addition, there is the volcanic North Sulawesi Magmatic Arc, comprising the remaining part of the north arm.

In the following, these tectonic provinces will be referred to as west-, south-east-, east-, and north Sulawesi, respectively. The island was formed in the accretion process of these four fragments to Sundaland from the Cretaceous onward, starting with the placement of west Sulawesi adjacent to south-east Borneo. Subsequently, south-east and east Sulawesi collided with west Sulawesi in the Miocene (20–15 Mya). Finally, the Banggai-Sula and Tukang Besi platforms moved close to their present position in the late Miocene or early Pliocene (8 Mya). The final juxtaposition of these fragments was achieved by internal rotation through a number of linked strike-slip faults and thrusts. While the scenario just described is apparently not very controversial, the distribution of land and sea on Sulawesi certainly is. According to Wilson & Moss (1999), Borneo and south Sulawesi were connected for a few million years in the early Eocene (c. 50 Mya). In addition, microcontinental fragments of the Tukang Besi and Banggai group might have been emerged during their passage from the Australian margin towards the rest of Sulawesi, thus suggesting two land migration routes to Sulawesi. Hall (1998, 2001), however, allows for the presence of land in south-east Sulawesi from the early Miocene (20 Mya) but maintains that west Sulawesi was submerged until the late Miocene (10 Mya), with rather small volcanic islands emerging at best. Towards Australia and the Philippines, island hopping might have been possible via volcanic island chains about 10–5 Mya.

The morphological data mentioned above and genetic data presented by Glaubrecht & Rintelen (2003) clearly indicate that the Sulawesi pachychilids do not have their origin in Sundaland. Thus, only one explanation for the present-day distribution of Tylomelania and Pseudopotamis consistent with the geological facts outlined above remains: a colonization of Sulawesi from the east on a terrane from the Australian margin (including New Guinea). However, several severe problems persist. One would expect from this hypothesis to find that the common ancestor of both taxa was widely distributed in Australia and/or New Guinea. However, no pachychilids are found east of Sulawesi and on the Australian mainland, even though extensive (if somehow random) sampling has taken place throughout the area, with the possible exception of the inner regions of New Guinea. Possibly, the extinction of pachychilids elsewhere in Australia (with the exception of the Torres Strait Islands), the result of drastic climatic changes in the region during the Pliocene, accounts for the observed pattern (see review in Glaubrecht & Rintelen, 2003). Consequently, we have suggested a relict status for Pseudopotamis. As the Torres Strait Islands were subject to the same climatic changes, however, the restricted distribution of Pseudopotamis on them remains enigmatic. A more critical obstacle is the lack of a geological scenario hypothesizing an actual, permantly afloat terrane covering the distance between the Australian margin and Sulawesi (see above). Assuming our biogeographical hypothesis is valid, there remains a conflict between the biological and geological data, which cannot be solved for the time being.

The alternative to vicariance – dispersal − is equally in conflict with the available data. For a hypothesis, one of two preconditions must be met. First, animals would have to have been dispersed across large stretches of ocean, which is only imaginable by some agent like birds and hardly testable. Beyond this general difficulty, recent dispersal, e.g. by humans, would be in conflict with the deep (ancient) split between Tylomelania and Pseudopotamis evident in the molecular phylogeny presented by Glaubrecht & Rintelen (2003). Alternatively, the striking similarity in the reproductive anatomy and biology of Tylomelania and Pseudopotamis could have developed convergently after their separation, i.e. their viviparity arose after the colonization of Sulawesi or the Torres Strait Islands. The common reproductive characters have so far been interpreted as homologous and synapomorphic for both taxa, and it does not seem parsimonious to assume a convergent evolution for both. Even if a more detailed study of the anatomy of Pseudopotamis could provide support for an independent origin of viviparity in both taxa, the biogeographical problem outlined above is not eliminated. A dispersal stage able to survive marine transport is still required, and there is no evidence for the existence of veliger larvae in pachychilids so far.

In conclusion, we favour an origin of Tylomelania following the separation of populations of an ovoviviparous pachychilid ancestor which also gave rise to Pseudopotamis, as such an hypothesis enjoys better support from both morphological and molecular data, and requires far fewer ad-hoc assumptions than a dispersal scenario. The biological data indicate the need to develop more detailed geological hypotheses on the movement of plate fragments and their submergence in the region east of Sulawesi.

ACKNOWLEDGEMENTS

We are very grateful to Ristiyanti Marwoto (MZB) for her support in arranging the field trips to Sulawesi, which helped make this study possible. We also thank LIPI (Indonesian Institute of Sciences) for permits to conduct research in Indonesia. Invaluable support was provided by the staff of INCO in Soroako, Lake Matano. Without their help in providing accommodation, transport and general logistics this study would not have been such a success.

Philippe Bouchet (MNHN) initiated the study of Sulawesi pachychilids by visiting the lakes and making all of his material available for a pilot study. Ambros Hänggi, Urs Wüest (NMB), Robert Moolenbeck (ZMA) and Trudi Meier (ZMZ) are thanked for their courtesy and generous help with the loan of material. Ellen Strong shared her anatomical insight. Nora Brinkmann was immensely helpful with radula preparation. Thanks also to M. Drescher, S. Schütt, J. Zeller (ZMB) and A. Munandar (MZB) for technical assistance. The comments by Ellinor Michel (BMNH) and one anonymous reviewer helped improve the manuscript. This study was financed by grants GL 297/1–1/−2, and GL 297/7–1 of the Deutsche Forschungsgemeinschaft (DFG).

REFERENCES

Adams
H
1866
.
Description of a new genus and a new species of molluscs
.
Proceedings of the Zoological Society of London
 
34
:
150
151
.
Barlow
GW
2000.
The cichlid fishes: nature's grand experiment in evolution
 .
Cambridge, MA
:
Perseus Publishing
.
Berry
AJ
,
Kadri
AB
1974
.
Reproduction in the Malayan freshwater cerithioidean gastropod Melanoides tuberculata
.
Journal of Zoology
 
172
:
369
381
.
Binder
E
1959
.
Anatomie et systématiques des melaniens d'Afrique occidentale
.
Revue Suisse de Zoologie
 
66
:
735
759
.
Boss
KJ
1978
.
On the evolution of gastropods in ancient lakes
. In:
Fretter
V
Peake
JF
eds.
Pulmonates
 ,
Vol. 2A
.Systematics, evolution and ecology.
London
:
Academic Press
,
385
428
.
Brooks
JL
1950
.
Speciation in ancient lakes
.
Quarterly Review of Biology
 
25
(
30–60
):
131
176
.
Brown
DS
1994.
Freshwater snails of Africa and their medical importance
 .
London
:
Taylor & Francis
.
Burns
KJ
,
Hackett
SJ
,
Klein
NK
2002
.
Phylogenetic relationships and morphological diversity in Darwin's finches and their relatives
.
Evolution
 
56
:
1240
1252
.
Calow
P
1978
.
The evolution of life-cycle strategies in fresh-water gastropods
.
Malacologia
 
17
:
351
364
.
Cohen
AS
,
Johnston
MR
1987
.
Speciation in brooding and poorly dispersing lacustrine organisms
.
Palaios
 
2
:
426
435
.
Davis
GM
1971
.
Systematic studies of Brotia costula episcopalis, first intermediate host of Paragonimus westermani in Malaysia
.
Proceedings of the Academy of Natural Sciences in Philadelphia
 
123
:
53
86
.
Davis
GM
1972
.
Geographic variation in Semisulcospira libertina (Mesogastropoda: Pleuroceridae)
.
Proceedings of the Malacological Society
 
40
:
5
32
.
Davis
GM
1982
.
Historical and ecological factors in the evolution, adaptive radiation, and biogeography of freshwater mollusks
.
American Zoologist
 
22
:
375
395
.
Dillon
RT
1984
.
What shall I measure on my snails? Allozyme data and multivariate analysis used to reduce the non-genetic component of morphological variance in Goniobasis proxima
.
Malacologia
 
25
:
503
511
.
Dillon
RT
2000.
The ecology of freshwater molluscs
 .
Cambridge
:
Cambridge University Press
.
Dudgeon
D
1982
.
The life history of Brotia hainanensis (Brot, 1872) (Gastropoda: Prosobranchia: Thiaridae) in a tropical forest stream
.
Zoological Journal of the Linnean Society
 
76
:
141
154
.
Dudgeon
D
1989
.
Ecological strategies of Hong Kong Thiaridae (Gastropoda: Prosobranchia)
.
Malacological Review
 
22
:
39
53
.
Dybdahl
MF
,
Lively
CM
1996
.
The geography of coevolution: comparative population structures for a snail and its trematode parasite
.
Evolution
 
50
:
2264
2275
.
Fischer
PH
,
Crosse
JCH
1892.
Pour servir a l'histoire de la faune de l'Amerique Centrale et du Mexique
 .
Paris
:
Imprimerie Nationale.
Fretter
V
,
Graham
A
1994.
British prosobranch molluscs
 .
Andover
:
The Ray Society.
Frömming
E
1956.
Biologie der Mitteleuropäischen Süßwasserschnecken
 .
Berlin
:
Duncker & Humblot.
Givnish
TJ
,
Sytsma
KJ.
eds.
1997.
Molecular evolution and adaptive radiation.
 
Cambridge
:
Cambridge University Press.
Glaubrecht
M
1996
.
Evolutionsökologie und Systematik am Beispiel von Süβ- und Brackwasserschnecken (Mollusca: Caenogastropoda: Cerithioidea): Ontogenese-Strategien, Paläontologische Befunde und Historische Zoogeographie
 .
Leiden
:
Backhuys
.
Glaubrecht
M
1999
.
Systematics and the evolution of viviparity in tropical freshwater gastropods (Cerithioidea: Thiaridae sensu lato) – an overview
.
Courier Forschungsinstitut Senckenberg
 
125
:
91
96
.
Glaubrecht
M
,
Rintelen
T von
2003
.
Systematics and zoogeography of the pachychilid gastropod Pseudopotamis Martens, 1894 (Mollusca: Gastropoda: Cerithioidea): a limnic relict on the Torres Strait Islands, Australia?
Zoologica Scripta
 
32
:
415
435
.
Glaubrecht
M
,
Strong
EE
2004
.
Spermatophores of thalassoid gastropods (Paludomidae) in Lake Tanganyika, East Africa, with a survey of their occurrence in Cerithioidea: functional and phylogenetic implications
.
Invertebrate Biology
 
123
:
218
236
.
Grant
PR
,
Grant
BR
2002
.
Adaptive radiation of Darwin's finches
.
American Scientist
 
90
:
130
139
.
Hall
R
1996.
Reconstructing Cenozoic SE Asia
 . In:
Hall
R
Blundell
DJ
eds.
Geological Society Special Publication
 , Vol.
106
.
Tectonic evolution of Southeast Asia.
 
London
:
Geological Society
,
153
184
.
Hall
R
1998
.
The plate tectonics of Cenozoic SE Asia and the distribution of land and sea
. In:
Hall
R
Holloway
JD
eds.
Biogeography and geological evolution of SE Asia
 .
Leiden
:
Backhuys
,
99
131
.
Hall
R
2001
.
Cenozoic reconstructions of SE Asia and the SW Pacific: changing patterns of land and sea
. In:
Metcalfe
I
Smith
JMB
Morwood
M
Davidson
I
eds.
Faunal and floral migrations and evolution in SE Asia-Australasia
 .
Lisse
:
Balkema
,
35
56
.
Holznagel
WE
1998
.
A nondestructive method for cleaning gastropod radulae from frozen, alcohol-fixed, or dried material
.
American Malacological Bulletin
 
14
:
181
183
.
Huelsenbeck
JP
,
Ronquist
F
2001a.
MrBayes: A program for the Bayesian inference of phylogeny
 . morphbank.ebc.uu.se/mrbayes
Huelsenbeck
JP
,
Ronquist
F
2001b
.
MRBAYES: Bayesian inference of phylogenetic trees
.
Bioinformatics
 
17
:
754
755
.
Huelsenbeck
JP
,
Ronquist
F
,
Nielsen
R
,
Bollback
JP
2001
.
Bayesian inference of phylogeny and its impact on evolutionary biology
.
Science
 
294
:
2310
2314
.
Johnson
SG
1994
.
Parasitism, reproductive assurance and the evolution of reproductive mode in a freshwater snail
.
Proceedings of the Royal Society of London B
 
255
:
209
213
.
Johnston
MR
,
Cohen
AS
1987
.
Morphological divergence in endemic gastropods from Lake Tanganyika: Implications for models of species flock evolution
.
Palaios
 
2
:
413
425
.
Köhler
F
2003.
Brotia in space and time – phylogeny and evolution of Southeast Asian freshwater gastropods of the family Pachychilidae (Caenogastropoda, Cerithioidea)
 .
Unpublished PhD Thesis
 , Humboldt-Universität zu Berlin.
Köhler
F
,
Glaubrecht
M
2001
.
Toward a systematic revision of the Southeast Asian freshwater gastropod Brotia H. Adams, 1866 (Cerithioidea: Pachychilidae): an account of species from around the South China Sea
.
Journal of Molluscan Studies
 
67
:
281
318
.
Köhler
F
,
Glaubrecht
M
2003
.
Morphology, reproductive biology and molecular genetics of ovoviviparous freshwater gastropods (Cerithioidea, Pachychilidae) from the Philippines, with description of a new genus Jagora
.
Zoologica Scripta
 
32
:
35
59
.
Köhler
F
,
Rintelen
T von
,
Glaubrecht
M
2000
.
Morphology, molecules and Wallacean biogeographic. The case study of the Southeast Asian freshwater gastropod Brotia (Cerithioidea: Pachychilidae)
. In:
Jong
Rd
ed.
Abstracts of the International Symposium Biogeography of SE Asia 2000.
 
Leiden
,
51
52
.
Kornfield
I
,
Smith
PF
2000
.
African cichlid fishes: model systems for evolutionary biology
.
Annual Review of Ecology and Systematics
 
31
:
163
196
.
Korniushin
AV
,
Glaubrecht
M
2003
.
Novel reproductive modes in freshwater clams: brooding and larval morphology in Southeast Asian taxa of Corbicula (Mollusca, Bivalvia, Corbiculidae)
.
Acta Zoologica
 
84
:
293
315
.
Kruimel
JH
1913
.
Verzeichnis der von Herrn E.C. Abendanon in Celebes gesammelten Süsswasser-Mollusken
.
Bijdragen Tot de Dierkunde
 
19
:
217
235
.
Lack
D
1947.
Darwin's finches
 .
Cambridge
:
Cambridge University Press
.
Lea
I
,
Lea
HC
1850
.
Description of a new genus of the family of Melaniana, and of many new species of the genus Melania, chiefly collected by Hugh Cuming, Esq., during his zoological voyage in the East, and now first described
.
Proceedings of the Zoological Society London
 
18
:
179
246
.
Lively
CM
,
Johnson
SG
1994
.
Brooding and the evolution of parthenogenesis: strategy models and evidence from aquatic invertebrates
.
Proceedings of the Royal Society of London B
 
256
:
89
95
.
Lydeard
C
,
Holznagel
WE
,
Glaubrecht
M
,
Ponder
WF
2002
.
Molecular phylogeny of a circum-global, diverse gastropod superfamily (Cerithioidea: Mollusca: Caenogastropoda): pushing the deepest phylogenetic limits of mitochondrial LSU rDNA sequences
.
Molecular Phylogenetics and Evolution
 
22
:
399
406
.
MacArthur
RH
,
Wilson
EO
1967.
The theory of island biogeography
 .
Princeton, NJ
:
Princeton University Press
.
Martens
E von
1894
.
Mollusken
.
Zoologische Forschungsreisen in Australien und deem Malayischen Archipel
 ,
5
. Systematik und Thiergeographie.
Jena
:
Gustav Fischer
,
82
88
.
Marwoto
R
1997
.
A preliminary study of the biodiversity of the freshwater snail family Thiaridae from Indonesia (Mollusca: Prosobranchia)
. In:
Ulrich
H
ed.
Proceedings of the International Symposium on Biodiversity and Systematics in Tropical Ecosystems. Tropical Biodiversity and Systematics.
 
Bonn
:
Zoologisches Forschungsinstitut und Museum Alexander Koenig
,
109
112
.
Metcalfe
I
1996.
Pre-Cretaceous evolution of SE Asian terranes
 . In:
Hall
R
Blundell
DJ
eds.
Geological Society Special Publicatio
 n Vol.
106
.
Tectonic evolution of Southeast Asia.
 
London
:
Geological Society
,
97
122
.
Metcalfe
I
1998
.
Paleozoic and Mesozoic geological evolution of the SE Asian region: multidisciplinary constraints and implications for biogeography
. In:
Hall
R
Holloway
JD
eds.
Biogeography and geological evolution of SE Asia
 .
Leiden
:
Backhuys
,
25
41
.
Metcalfe
I
2001
.
Paleozoic and Mesozoic tectonic evolution and biogeography of SE Asia-Australasia
. In:
Metcalfe
I
Smith
JMB
Morwood
M
Davidson
I
eds.
Faunal and floral migrations and evolution in SE Asia-Australasia
 .
Lisse
:
Balkema
,
15
34
.
Meyer
A
1993
.
Phylogenetic relationships and evolutionary processes in East African cichlid fishes
.
Trends in Ecology and Evolution
 
8
:
279
284
.
Meyer
A
,
Lydeard
C
1993
.
The evolution of copulatory organs, internal fertilization, placentae and viviparity in killifishes (Cyprinodontiformes) inferred from a DNA phylogeny of the tyrosine kinase gene X-src
.
Proceedings of the Royal Society of London B
 
254
:
153
162
.
Michel
E
1994
.
Why snails radiate: a review of gastropod evolution in long-lived lakes, both recent and fossil
 . In:
Martens
K
Goddeeris
B
Coulter
G
eds, Vol.
44
.
Speciation in ancient lakes.
 
Stuttgart
:
E. Schweitzerbart'sche-Verlagsbuchhandlung
,
285
317
.
Michel
E
2000
.
Phylogeny of a gastropod species flock: exploring speciation in Lake Tanganyika in a molecular framework
. In:
Rossiter
A
Kawanabe
H
eds.
Advances in ecological research
 , Vol.
31
.Ancient lakes: biodiversity, ecology and evolution.
SanDiego
:
Academic Press
,
275
302
.
Morrison
JPE
1954
.
The relationships of old and new world melanians
.
Proceedings of the United States National Museum
 
103
:
357
394
.
Moss
SJ
,
Wilson
EJ
1998
.
Biogeographic implications of the Tertiary palaeogeographic evolution of Sulawesi and Borneo
. In:
Hall
R
Holloway
JD
eds.
Biogeography and geological evolution of SE Asia
 .
Leiden
:
Backhuys Publishers
,
133
163
.
Mousson
A
1849.
Die Land- und Süßwasser-Mollusken von Java
 .
Zürich
:
Friedrich Schulthess.
Nylander
JAA
2002.
Mr Modeltest
  V1.0b.
Uppsala
:
Department of Systematic Zoology, Uppsala University
.
Posada
D
,
Crandall
KA
2001
.
Selecting the best-fit model of nucleotide substitution
.
Systematic Biology
 
50
:
580
501
.
Reid
DG
1996.
Systematics and evolution of Littorina
 .
Andover
:
The Ray Society.
Riech
E
1937
.
Systematische, anatomische, ökologische und tiergeographische Untersuchungen über die Süßwassermollusken Papuasiens und Melanesiens
.
Archiv für Naturgeschichte
 
6
:
37
153
.
Rintelen
T von
2003.
Phylogenetic analysis and systematic revision of a species flock of viviparous freshwater gastropods in the ancient lakes on Sulawesi, Indonesia – a model case of adaptive radiation?
 
Unpublished PhD Thesis
 , Humboldt-Universität zu Berlin.
Rintelen
T von
,
Glaubrecht
M
1999
.
On the reproductive anatomy of freshwater gastropods of the genera Brotia H. Adams, 1866 and Tylomelania Sarasin & Sarasin, 1897 in the central lakes on Sulawesi, Indonesia (Cerithioidea: Melanatriidae)
.
Courier Forschungsinstitut Senckenberg
 
125
:
163
170
.
Rintelen
T von
,
Glaubrecht
M
2003
.
New discoveries in old lakes: three new species of Tylomelania Sarasin & Sarasin, 1897 (Gastropoda: Cerithioidea: Pachychilidae) from the Malili lake system on Sulawesi, Indonesia
.
Journal of Molluscan Studies
 
69
:
3
17
.
Rintelen
T von
,
Wilson
AB
,
Meyer
A
&
Glaubrecht
M
2004.
Escalation and trophic specialization drive adaptive radiation of viviparous freshwater gastropods in the ancient lakes on Sulawesi, Indonesia
 .
Proceedings of the Royal Society of London B
 
271
:
2541
2549
.
Salzburger
W
,
Meyer
A
,
Baric
S
,
Verheyen
E
,
Sturmbauer
C
2002
.
Phylogeny of the Lake Tanganyika cichlid species flock and its relationship to the Central and East African haplochromine cichlid fish faunas
.
Systematic Biology
 
51
:
113
135
.
Sarasin
P
,
Sarasin
F
1897
.
Über die Molluskenfauna der großen Süßwasser-Seen von Central-Celebes
.
Zoologischer Anzeiger
 
539/540
:
308
320
.
Sarasin
P
,
Sarasin
F
1898.
Die Süßwassermollusken von Celebes
 .
Wiesbaden
:
Kreidel.
Sarasin
P
,
Sarasin
F
1905.
Reisen in Celebes Ausgeführt in den Jahren 1893–1896 und 1902–1903
 .
Wiesbaden
:
Kreidel.
Schluter
D
2000.
The ecology of adaptive radiation
 .
Oxford
:
Oxford University Press
.
Sherbakov
DY
1999
.
Molecular phylogenetic studies on the origin of biodiversity in Lake Baikal
.
Trends in Ecology and Evolution
 
14
:
92
95
.
Simone
LRL
2001
.
Phylogenetic analyses of Cerithioidea (Mollusca, Caenogastropoda) based on comparative morphology
.
Arqivos de Zoologia
 
36
:
147
263
.
Solem
A
1966
.
Some non-marine mollusks from Thailand, with notes on classification of the Helicarionidae
.
Spolia Zoologica Musei Hauniensis
 
24
:
1
110
.
Starmühlner
F
1969
.
Die Gastropoden der Madagassischen Binnengewässer
.
Malacologia
 
8
:
1
434
.
Strong
EE
,
Glaubrecht
M
2002
.
Evidence for convergent evolution of brooding in a unique gastropod from Lake Tanganyika: anatomy and affinity of Tanganyicia rufofilosa (Caenogastropod, Cerithioidea, Paludomidae)
.
Zoologica Scripta
 
31
:
167
184
.
Strong
EE
,
Glaubrecht
M
2003
.
Anatomy and systematic affinity of Stanleya neritinoides (Smith, 1880), an enigmatic member of the thalassoid gastropod species flock in Lake Tanganyika, East Africa (Cerithioidea, Paludomidae)
.
Acta Zoologica
 
84
:
249
265
.
Sturmbauer
C
1998
.
Explosive speciation in cichlid fishes of the African Great Lakes: a dynamic model of adaptive radiation
.
Journal of Fish Biology
 
53
:
18
36
.
Sturmbauer
C
,
Meyer
A
1992
.
Genetic divergence, speciation and morphological stasis in a lineage of African cichlid fishes
.
Nature
 
358
:
578
581
.
Swofford
DL
1998.
PAUP*: phylogenetic analysis using parsimony (*and other methods)
  Version 4.0.Sunderland,
MA
:
Sinauer
.
Takami
A
1991
.
The birth frequency, number and size of newborns in the three species of the genus Semisulcospira (Prosobranchia: Pleuroceridae)
.
Venus
 
50
:
218
232
.
Thiele
J
1921
.
Zur Systematik der Mollusken
.
Archiv für Molluskenkunde
 
53
:
140
163
.
Thiele
J
1925.
Mollusca
 . In:
Kükenthal
W
ed. Handbuch der Zoologie, Vol.
5
.
Leipzig
:
Metzger & Wittig
,
1
256
.
Thiele
J
1928
.
Revision des Systems der Hydrobiiden und Melaniiden
.
Zoologische Jahrbücher
 
55
:
351
402
.
Thiele
J
1929.
Handbuch der Systematischen Weichtierkunde
 .
Jena
:
Gustav Fischer
.
Thompson
JD
,
Gibson
TJ
,
Plewniak
F
,
Jeanmougin
F
,
Higgins
DG
1997
.
The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools
.
Nucleic Acids Research
 
25
:
4876
4882
.
Troschel
FH
1857.
Das Gebiss der Schnecken zur Begründung einer Natürlichen Classification
 .
Berlin
:
Nicolaische Verlagsbuchhandlung
.
Wesenberg-Lund
C
1939.
Biologie der Süsswassertiere. Wirbellose Tiere
 .
Wien
:
Julius Springer
.
West
K
,
Michel
E
2000
.
The dynamics of endemic diversification: Molecular phylogeny suggests an explosive origin of the thiarid gastropods of Lake Tanganyika
. In:
Rossiter
A
Kawanabe
H
eds.
Advances in ecological research
 , Vol.
31
Ancient lakes: biodiversity, ecology and evolution.
SanDiego
:
Academic Press
,
331
354
.
West
K
,
Michel
E
,
Todd
J
,
Brown
D
,
Clabaugh
J
2003.
The gastropods of Lake Tanganyika
 . Chapel
Hill, NC
:
International Association of Theoretical and Applied Limnology.
Wilson
AB
,
Glaubrecht
M
,
Meyer
A
2004
.
Ancient lakes as evolutionary reservoirs: evidence from the thalassoid gastropods of Lake Tanganyika
.
Proceedings of the Royal Society of London B
 
271
:
529
536
.
Wilson
MEJ
,
Moss
SJ
1999
.
Cenozoic palaeogeographic evolution of Sulawesi and Borneo
.
Palaeogeography, Palaeoclimatology, Palaeoecology
 
145
:
303
337
.
Winnepenninckx
B
,
Backeljau
T
,
De Wachter
R
1993
.
Extraction of high molecular weight DNA from molluscs
.
Trends in Genetics
 
9
:
407
.

APPENDIX

Museum voucher and GenBank numbers of sequenced specimens.

Species Museum voucher no. GenBank accession no. Source 
Pseudopotamis supralirata ZMB 190363 AY242970 Glaubrecht & Rintelen (2003
Tylomelania carbo ZMB 190200 AY311825 this study 
T. centaurus ZMB 190021 AY311830 this study 
T. connectens ZMB 190017 AY311826 this study 
T. gemmifera ZMB 190051 AY242954 Glaubrecht & Rintelen (2003
T. insulaesacrae ZMB 190122 AY311858 this study 
T. kruimeli ZMB 190155 AY311852 this study 
T. kuli ZMB 190011 AY311839 this study 
T. neritiformis ZMB 190016 AY242960 Glaubrecht & Rintelen (2003
T. patriarchalis ZMB 190087 AY311871 this study 
T. perfecta ZMB 190086 AY242958 Glaubrecht & Rintelen (2003
T. sarasinorum ZMB 190212 AY311906 this study 
T. toradjarum ZMB 190203 AY311836 this study 
T. towutensis ZMB 190115 AY311912 this study 
T. towutica ZMB 190116 AY311924 this study 
T. zeamais ZMB 190052 AY311931 this study 
Species Museum voucher no. GenBank accession no. Source 
Pseudopotamis supralirata ZMB 190363 AY242970 Glaubrecht & Rintelen (2003
Tylomelania carbo ZMB 190200 AY311825 this study 
T. centaurus ZMB 190021 AY311830 this study 
T. connectens ZMB 190017 AY311826 this study 
T. gemmifera ZMB 190051 AY242954 Glaubrecht & Rintelen (2003
T. insulaesacrae ZMB 190122 AY311858 this study 
T. kruimeli ZMB 190155 AY311852 this study 
T. kuli ZMB 190011 AY311839 this study 
T. neritiformis ZMB 190016 AY242960 Glaubrecht & Rintelen (2003
T. patriarchalis ZMB 190087 AY311871 this study 
T. perfecta ZMB 190086 AY242958 Glaubrecht & Rintelen (2003
T. sarasinorum ZMB 190212 AY311906 this study 
T. toradjarum ZMB 190203 AY311836 this study 
T. towutensis ZMB 190115 AY311912 this study 
T. towutica ZMB 190116 AY311924 this study 
T. zeamais ZMB 190052 AY311931 this study